U.S. patent application number 15/424976 was filed with the patent office on 2017-05-25 for methods and systems for determining information relating to the operation of traffic control signals.
The applicant listed for this patent is TomTom International B.V. Invention is credited to Stefan Nico Anton Bollars, Paul Gerard Krijger.
Application Number | 20170148314 15/424976 |
Document ID | / |
Family ID | 49033556 |
Filed Date | 2017-05-25 |
United States Patent
Application |
20170148314 |
Kind Code |
A1 |
Krijger; Paul Gerard ; et
al. |
May 25, 2017 |
METHODS AND SYSTEMS FOR DETERMINING INFORMATION RELATING TO THE
OPERATION OF TRAFFIC CONTROL SIGNALS
Abstract
Data indicative of the durations of multiple instances of
different phases of a traffic control signal in a given time period
is determined. The data is used to obtain data indicative of a
distribution of the durations of each phase. The distribution data
is used to obtain data indicative of a probability of the traffic
control signal having a given phase at one or more future time. The
probability data may be used to provide an expected waiting time
when arriving at the signal at a future time and/or a speed
recommendation for a vehicle approaching the signal.
Inventors: |
Krijger; Paul Gerard;
(Utrecht, NL) ; Bollars; Stefan Nico Anton;
(Helmond, NL) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
TomTom International B.V |
Amsterdam |
|
NL |
|
|
Family ID: |
49033556 |
Appl. No.: |
15/424976 |
Filed: |
February 6, 2017 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14326535 |
Jul 9, 2014 |
9564050 |
|
|
15424976 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G08G 1/0129 20130101;
G08G 1/082 20130101; G08G 1/096827 20130101; G08G 1/096741
20130101; G08G 1/096 20130101; G08G 1/096844 20130101; G01C 21/3492
20130101; G08G 1/0145 20130101; G08G 1/096716 20130101 |
International
Class: |
G08G 1/01 20060101
G08G001/01; G08G 1/082 20060101 G08G001/082 |
Foreign Application Data
Date |
Code |
Application Number |
Jul 9, 2013 |
GB |
1312306.2 |
Claims
1. A method for determining information relating to the operation
of at least one traffic control signal, the traffic control signal
being operable to transition between different phases in use, the
method comprising: using data indicative of the durations of
multiple instances of at least one phase of the traffic control
signal to determine, for the or each phase, data indicative of a
distribution of the durations of the multiple instances of the
phase; and using the determined distribution data to obtain data
indicative of a probability of the traffic control signal having a
given phase at one or more future time.
2. The method of claim 1, wherein the data indicative of the
durations of multiple instances of at least one phase of the
traffic control signal is indicative of the durations of multiple
instances of the at least one phase of the traffic control signal
in a given time period, and the data indicative of the distribution
of the durations of the multiple instances of the or each phase is
indicative of the durations of the multiple instances of the or
each phase in at least a portion of the given time period.
3. The method of claim 1, further comprising obtaining the data
indicative of the durations of multiple instances of the at least
one phase of the traffic control signal from one or more of: a
third party information, a server, the traffic control signal, a
vehicle, and positional data indicative of the movement of a
plurality of devices with respect to time along a path controlled
by the traffic control signal.
4. The method of claim 1, wherein the at least one phase of the
traffic control signal for which duration data is determined
comprises a phase having a duration that is variable in response to
demand.
5. The method of claim 1, wherein the probability data is obtained
by determining the probability that each of said one or more future
times falls within said given phase of said traffic control
signal.
6. The method of claim 1, wherein the step of determining the
probability data comprises determining the probability of the or
each future time coinciding with the given phase for each of a
plurality of possible cycle plans of the traffic control signal,
and combining the probabilities for each possible cycle plan.
7. The method of claim 1, further comprising using data indicative
of a timing of at least one instance of a phase of the traffic
control signal in the given time period to which the duration data
relates together with the duration data in obtaining the
probability data.
8. The method of claim 1, wherein the probability data is
indicative of the probability of the traffic control signal having
the given phase with respect to time over a given future time
period.
9. The method of claim 1 wherein the given phase to which the
probability data relates is a phase permitting the flow of traffic
along a path controlled by the signal, preferably wherein the phase
has a duration that is variable in response to demand.
10. The method of claim 1, further comprising identifying data
indicative of one or more maximum in the probability data with
respect to time, and determining a time associated with the or each
maximum.
11. The method of claim 1, further comprising using the determined
probability data to provide a speed recommendation for a vehicle to
enable the vehicle to arrive at the traffic control signal at or
around a time which is expected to coincide with a phase of the
signal allowing the passage of traffic along a path controlled by
the traffic control signal based on the probability data.
12. The method of claim 1 further comprising using the obtained
probability data indicative of the probability of the traffic
control signal having a given phase at a future time or times to
determine an expected waiting time for a vehicle when arriving at
the signal at one or more future time of interest.
13. The method of claim 12, wherein the future time of interest is
a time at which the vehicle is expected to arrive at the traffic
control signal when following a given route, the method further
comprising using the determined expected waiting time in
determining an estimated travel time for the route.
14. The method of claim 12, comprising obtaining data indicative of
expected waiting time with respect to time of arrival at the
traffic control signal at different times in a given future time
period, and using the data to determine a speed recommendation for
a vehicle, wherein the speed recommendation is a speed
recommendation that is expected to minimise expected waiting time
at the signal.
15. The method of claim 12, wherein the or each traffic control
signal is associated with navigable segments of a navigable
network, the method comprising obtaining data indicative of an
expected waiting time in respect of the or each traffic control
signal and using the expected waiting time data in generating a
route to a destination through the navigable network, optionally
that minimises expected waiting time at traffic control signals
along the route.
16. A system for determining information relating to the operation
of at least one traffic control signal, the traffic control signal
being operable to transition between different phases in use, the
system comprising at least one processor and at least one memory
including computer program code, the at least one memory and the
computer program code configured to, with the at least one
processor, cause the apparatus to at least: use data indicative of
the durations of multiple instances of at least one phase of the
traffic control signal to determine, for the or each phase, data
indicative of a distribution of the durations of the multiple
instances of the phase; and use the determined distribution data to
obtain data indicative of a probability of the traffic control
signal having a given phase at one or more future time.
17. A non-transitory computer readable medium comprising computer
program code that, when executed on a computer, cause the computer
to perform a method according claim 1.
Description
CROSS-REFERENCES TO RELATED APPLICATIONS
[0001] The present application is a continuation of U.S.
application Ser. No. 14/326,535, filed Jul. 9, 2014, and further
claims priority from United Kingdom Patent Application No.
1312306.2, filed Jul. 9, 2013. The entire content of these
applications is incorporated herein by reference.
FIELD OF THE INVENTION
[0002] The present invention relates to methods and systems for
determining information relating to the operation of traffic
control signals, and in particular, although not exclusively, to
methods for predicting the timing of future phases of a traffic
control signal. In some aspects and embodiments, the present
invention extends to methods for determining expected waiting times
at traffic control signals. The present invention also extends to
methods of using determined information relating to the operation
of traffic control signals. In accordance with yet further aspects
and embodiments, the present invention relates to methods of
determining a route through a network of navigable segments and/or
in providing speed recommendations.
BACKGROUND OF THE INVENTION
[0003] Information relating to the operation of traffic control
signals may be useful in various contexts. Methods and systems have
previously been proposed in which information or recommendations
are provided to drivers based upon information relating to the
operation of traffic control signals, e.g. traffic lights. In some
methods, information may be provided to drivers regarding the state
of upcoming traffic control signals, e.g. along a route being
navigated. The information may be used to provide a speed
recommendation to drivers. For example, a driver may be advised as
to an appropriate speed of travel to enable them to arrive at a
traffic control signal in order to coincide with a green phase of
the signal, i.e. to ride a "green wave" through a series of traffic
control signals. Information about the operation of traffic control
signals may be used to advise as to appropriate speeds of travel to
enable a driver to travel through a region containing one or more
sets of traffic control signals in a more efficient manner, in
terms of travel time and/or fuel usage. Knowledge of traffic
control signal operation is also useful in determining more
accurate travel times, e.g. by navigation devices, or for
infrastructure planning, etc, and in optimising routes, e.g. with
respect to travel time.
[0004] Information about the operation of traffic control signals
has previously often relied upon traffic control signal operation
data obtained from third party sources (e.g. governmental traffic
sources). Such data may often be based upon data collected from
fixed traffic sensors in the vicinity of traffic control signals.
Techniques of this type offer limited flexibility in terms of the
data available and the traffic control signals for which data is
provided, and are relatively expensive to implement, requiring the
appropriate fixed infrastructure to be in place.
[0005] WO 2013/060774 A1 entitled "Methods and Systems for
determining information relating to the operation of traffic
control signals" describes methods of determining information
relating to the operation of a traffic control signal using
positional data relating to the movement of vehicles with respect
to time along a path controlled by the traffic control signal
(so-called vehicle "probe" data). The application describes methods
by which the probe data may be used to predict future transition
times between phases of the traffic signal. A cycle time for the
signal may be derived.
[0006] The Applicant has realised that there remains a need for
further methods and systems for determining information relating to
the operation of traffic control signals, and which, in particular,
although not exclusively, may be used to determine information
relating to the operation of traffic control signals which do not
operate in accordance with a predetermined cycle plan having phase
durations that are set in advance. Such traffic control signals may
be referred to as "dynamically managed" traffic control signals,
and operate in accordance with cycles in which the durations of
different phases are variable, usually between predetermined upper
and lower limits. The durations of the phases may typically be
variable in response to demand, e.g. based on the actual traffic
conditions at an intersection where the signal is located. For
example, where there is a large quantity of traffic wishing to
follow a particular path through the signal, the duration of a
phase of the traffic control signal permitting traffic flow along
that path may be increased to permit the passage of a greater
number of vehicles per cycle. Traffic control signals may be
dynamically managed in various manners. For example, the traffic
control signals themselves may be arranged to sense demand on the
approach to the signal, e.g. queuing traffic, or alternatively or
additionally traffic control signals may communicate directly with
approaching vehicles to sense their presence. In other
arrangements, traffic control signals may alternatively or
additionally be managed remotely through communication with a
traffic management centre, e.g. through wireless or wired
infrastructure, in a manner responsive to demand. It will be
appreciated that determining information relating to the operation
of such dynamically managed traffic control signals may present
particular challenges due to the inherent unpredictability of their
operation.
SUMMARY OF THE INVENTION
[0007] In accordance with a first aspect of the invention there is
provided a method for determining information relating to the
operation of at least one traffic control signal, the traffic
control signal being operable to transition between different
phases in use, the method comprising:
[0008] using data indicative of the durations of multiple instances
of at least one phase of the traffic control signal to determine,
for the or each phase, data indicative of a distribution of the
durations of the multiple instances of the phase; and
[0009] using the determined distribution data to obtain data
indicative of a probability of the traffic control signal having a
given phase at one or more future time.
[0010] Thus, in accordance with the invention, data relating to the
durations of multiple instances of one or more phase of a traffic
control signal is used to obtain, for the or each phase, data
indicative of a distribution of the durations of the multiple
instances of the phase of the signal. The distribution data is used
to determine a probability that the traffic control signal will
have a given phase at a particular future time of interest. Thus,
the invention uses probabilistic techniques to determine the
likelihood of the signal having a particular phase at a particular
future time based upon the actual operation of the signal.
[0011] In this way, the methods may be applied even to a traffic
control signal which has at least one phase whose duration is
variable in response to demand, i.e. which duration is not set in
advance. By using data indicative of durations of multiple
different instances of one or more phase of such a traffic signal
in operation, it is possible to predict future phase timings for
the signal using a probability based method. The invention does not
rely upon being able to determine fixed cycle time or predictable
timing of transitions between phases within a cycle. Of course,
while the invention is particularly applicable to traffic control
signals having at least one phase that is variable in response to
demand, the invention is not limited in its application to such
traffic control signals, and may equally be applied to traffic
control signals in which the durations of each phase are
predetermined, e.g. which operate in accordance with a
predetermined cycle plan. As the methods of the present invention
are based upon data indicative of the actual duration of phases of
a traffic control signal, they may be carried out using
appropriately obtained data, without needing to know whether or not
the phase durations of the signal are variable, and are able to
given an appropriate output regardless of whether one or more of
the phases of the signal turn out to be variable.
[0012] The present invention also extends to a system for
determining information relating to the operation of a traffic
control signal. Thus, in accordance with a second aspect of the
present invention there is provided a system, which may be a
server, for determining information relating to the operation of at
least one traffic control signal, the system comprising:
[0013] means for using data indicative of the durations of multiple
instances of at least one phase of the traffic control signal to
determine, for the or each phase, data indicative of a distribution
of the durations of the multiple instances of the phase; and
[0014] means for using the determined distribution data to obtain
data indicative of a probability of the traffic control signal
having a given phase at one or more future time.
[0015] The present invention in this further aspect may include any
or all of the features described in relation to the first aspect of
the invention, and vice versa, to the extent that they are not
mutually inconsistent. Thus, if not explicitly stated herein, the
system of the present invention may comprise means for carrying out
any of the steps of the method described.
[0016] The present invention is a computer implemented invention,
and any of the steps described in relation to any of the aspects or
embodiments of the invention may be carried out under the control
of a set of one or more processors. The means for carrying out any
of the steps described in relation to the system may be a set of
one or more processors. A given step may be carried out using the
same or a different set of processors to any other step. Any given
step may be carried out using a combination of sets of processors.
The system may further comprise data storage means, such as
computer memory, for storing, for example, the duration data, the
distribution data and/or the data indicative of the probability of
the traffic control signal having a given phase at one or more
future time.
[0017] In general, the system of the present invention in any of
its embodiments may be at least one processing device. The or a
processing device may be a device of a mobile device, such as a
navigation device, whether a PND or an integrated device, or may be
a device of a server.
[0018] In some embodiments the method of the present invention in
any of its aspects or embodiments is carried out using a navigation
device, and the present invention extends to a navigation device
arranged to carry out the steps of the method of any of the aspects
or embodiments of the invention. The navigation device may be a PND
or an integrated, e.g. in-vehicle, device. In accordance with any
of the aspects or embodiments of the invention the navigation
device may comprise a display for displaying an electronic map to a
user, a set of one or more processors configured to access digital
map data and cause an electronic map to be displayed to a user via
the display, and a user interface operable by a user to enable the
user to interact with the device. Thus, the system of the present
invention may be a navigation device, e.g. a processing device
thereof.
[0019] In other embodiments the method of the present invention in
any of its aspects or embodiments may be carried out by a server,
and the present invention extends to a server arranged to carry out
the steps of the method of any of the aspects or embodiments of the
invention. The system of the present invention of any of its
aspects or embodiments may be a server, e.g. a processing device
thereof.
[0020] Of course, the steps of the method of the present invention
in any of its aspects or embodiments may be carried out in part by
a server and in part by a navigation apparatus. The steps of the
method may be performed exclusively on a server, or some on a
server and the others on a navigation device in any combination, or
exclusively on a navigation device. Performance of one or more of
the steps on the server may be efficient and may reduce the
computational burden placed on a navigation device. Alternatively
if one or more steps are performed on the navigation device, this
may reduce any bandwidth required for network communication. Thus,
the system of the present invention may be provided in part by a
navigation device or other mobile device, and in part by a
server.
[0021] As used herein, the "duration" of a phase of the traffic
signal refers to the time that the traffic signal has that phase
for a particular instance of the phase. The duration of the phase
will be defined between a transition time from another phase to
that phase and a transition time from the phase to another phase.
As used herein, a transition time of the traffic control signal
refers to a time at which a transition between different phases of
the traffic control signal occurs. An instance of a phase refers to
a single occurrence of that phase, e.g. within a cycle of the
traffic control signal.
[0022] References to the "duration data" herein refer to the data
indicative of the durations of multiple instances of at least one
phase of the traffic control signal. References to the
"distribution data" herein refer to the data, for the or each
phase, indicative of the distribution of the durations of the
instances of the phase.
[0023] The methods of the present invention may be implemented in
relation to one or more traffic control signal. Thus, any of the
steps described herein in relation to "a traffic control signal"
may be carried out in relation to the or each traffic control
signal that is considered. For ease of reference the methods and
systems of the invention may be described in relation to "the" or
"a" traffic control signal. However, it will be appreciated that
the steps may equally applied to any or each other traffic control
signal considered where multiple traffic signals are involved.
[0024] The present invention relates to at least one traffic
control signal that is operable to transition between different
phases in use. The phases include a phase allowing traffic flow
along a path controlled by the traffic control signal and a phase
preventing traffic flow along a path controlled by the traffic
control signal. In other words the phases are "go" and "stop"
phases for the path being controlled. In preferred embodiments in
which the traffic control signal is a traffic light, the phases may
be red and green phases of the traffic light. Of course, the
traffic control signal cycle may, and typically does, comprise one
or more additional phases. In embodiments the traffic control
signal cycle further comprises a yellow phase. Such additional
phases, e.g. a yellow phase, may be considered as part of another
one of the phases, e.g. the red or green phase, for the purposes of
implementing the present invention, e.g. when identifying the
phase(s) for which distribution data is obtained. The appropriate
choice may depend upon the local law applying to the traffic
control signal. For example, where vehicles may travel along the
path controlled by the traffic control signal during a yellow
phase, such a phase may suitable be treated as a part of the green
phase. Alternatively additional phases may be treated as distinct
further phases such that the traffic control signal transitions
between three or more phases.
[0025] The traffic control signal may provide an indication of
phases other than in terms of a colour. For example, the phases may
be indicated by one or more symbols. Such arrangements may be used
in connection with traffic control signals for controlling
movements of public transport vehicles, e.g. trams, trains,
etc.
[0026] In accordance with the invention in any of its aspects or
embodiments, the traffic control signal is a traffic control signal
which may act to control different vehicle movements along a path.
The path may comprise at least a portion of one or more navigable
segments, e.g. at an intersection. The traffic control signal may
be any automated traffic control signal. Preferably the traffic
control signal is a traffic light. The traffic control signal is
preferably located at an intersection. The intersection is an
intersection where there are competing movements of traffic. The
intersection may be a roundabout, crossing or any type of
intersection. The traffic control signal may be one of a plurality
of traffic control signals located at the intersection.
[0027] The method of the invention involves using data indicative
of the durations of multiple instances of at least one phase of the
traffic control signal to determine distribution data which, in
accordance with the first and second aspects of the invention at
least, is then used to obtain probability data. It will be
appreciated that distribution data may be determined for only one
of the phases of the traffic control signal. However, in other
embodiments distribution data is obtained for two or more different
phases of the traffic control signal, and in preferred embodiments
for two different phases of the traffic control signal. The
distribution data is therefore obtained for at least some of the
phases of the traffic control signal, and in embodiments, each
phase of the traffic control signal. In embodiments the
distribution data is obtained at least for a phase allowing traffic
to pass along a path controlled by the traffic control signal and
optionally for a phase that prevents traffic from passing along the
path controlled by the traffic control signal. Where distribution
data is determined in relation to more than one phase of the
traffic control signal, this may be carried out in accordance with
any of the embodiments described. Thus, any discussion relating to
determining distribution data based thereon in relation to a phase,
may be understood to be applicable to such steps carried out in
relation to the or each of the at least one phase in respect of
which such determinations are made.
[0028] The traffic control signal operates in accordance with a
cycle comprising the different phases. Thus the phases of the
traffic control signal are phases of a cycle of the traffic control
signal. A given cycle of the traffic control signal is a cycle
containing a complete set of the different phases of the traffic
control signal through which the signal transitions. The sequence
of phases of the traffic control signal will be repeated in
subsequent cycles. Thus a given phase will have multiple instances
over time, over different cycles. The cycle length for a given
cycle is then defined by the time required to transition between a
complete sequence of phases of the traffic control signal. Each
transition between phases of the traffic control signal occurs at a
transition time for that particular transition. As discussed below,
a reference to the traffic control signal operating in accordance
with a cycle does not imply that the phase durations within the
cycle, or the timings of the phase transitions, are necessarily set
in advance.
[0029] The traffic control signal may transition between the
different phases in accordance with a predetermined automated
traffic control signal cycle plan. A cycle plan refers to the
operation of the traffic control signal over a plurality of cycles.
The cycle plan determines the cycle length of the traffic control
signal and the phases within the cycles e.g. the timing and/or
durations of the phases. A predetermined cycle plan is a cycle plan
in which the phase durations within each of a plurality of cycles
of the traffic control signal, and a length of each of the cycles,
are set in advance. Such a cycle plan will have a predetermined
cycle length or lengths. A traffic control signal of this type may
be referred to as a "statically managed" traffic control signal.
For traffic control signals operating in accordance with a
predetermined cycle plan, the duration of the phases within a
cycle, and hence the cycle length of the signal, may be time
dependent. For example, the cycle length may vary over the course
of a day. The traffic control signal may be controlled to operate
in accordance with one or more different predetermined cycle
lengths in different time periods, e.g. within a 24 hour period, on
different days of the week (such as the weekend versus weekdays),
at peak times and non peak times on particular days, etc. Thus,
phase durations of a traffic signal that operates in accordance
with a predetermined cycle plan, and the cycle times thereof, may
be fixed at least over a certain time period. While the phase
durations and cycle length may vary depending upon time within a
predetermined cycle plan, they vary in accordance with a cycle plan
that is set in advance, e.g. being pre-programmed, and not in
response to demand, i.e. as a result of actual traffic
conditions.
[0030] As discussed above, the invention is, however, particularly
applicable to traffic control signals which do not operate in
accordance with a predetermined automated traffic control signal
cycle plan, i.e. in which at least some phase durations within
cycles of the signal are responsive to demand, and are not set in
advance. Such traffic control signals do not have a predetermined
cycle length, and may be referred to as "dynamically managed". This
provides the ability to, for example, increase the "go" phase of
the traffic signal where there is little traffic, even if this is
in a period when heavier traffic might usually be expected (when a
"statically managed" traffic control signal might be operated in
accordance with a shorter predetermined "go" phase). It will be
appreciated that the traffic control signal may not then
necessarily be operated such that every incidence of a particular
phase is not set in advance. For example, a fixed phase length may
occasionally be used e.g. during initialization. However, in normal
operation, the traffic control signal is preferably operable such
that the duration of at least one, and preferably a plurality of
the different phases of the signal is variable in response to
demand.
[0031] A phase may have a duration that is variable in response to
demand within a predetermined permitted range of duration for that
phase. It will be appreciated that only one of the different phases
of the traffic control signal may be variable in this manner. In
such cases, the phase is preferably at least a phase that permits
flow of traffic along the path controlled by the signal, e.g. a
green phase, is variable in response to demand. Preferably a phase
that prevents flow of traffic along the path, e.g. a red phase, is
also variable in response to demand.
[0032] In preferred embodiments, the duration of one or more, and
preferably a plurality of, or each phase, of the traffic control
signal, is variable in response to demand. Thus at least some of
the phases of the traffic control signal have a duration that is
variable in response to demand. The demand is preferably a
vehicular demand. Thus, the operation of the traffic control signal
is based upon demand by vehicles, and not just pedestrians. In
accordance with preferred embodiments of the invention, the at
least one phase for which distribution data is obtained (and, in
accordance with the first and second aspects at least, used in
obtaining probability data) includes a phase that is variable in
response to demand (e.g. vehicular demand), and, in embodiments,
the or each of the at least one phase is variable in response to
demand. In some embodiments the traffic control signal may comprise
a further phase which has a duration which is not responsive to
demand, i.e. which is of fixed duration. In some embodiments at
least a phase of the traffic control signal that allows traffic to
pass along a path controlled by the traffic control signal is
variable in response to demand, and distribution data is obtained
at least for such a phase of the traffic control signal. The phase
that is variable in response to demand may be a green phase of the
signal.
[0033] It will be appreciated that the present invention may be
applied to a traffic control signal where it is uncertain whether
one or more phases thereof are of variable duration. The methods of
the invention may equally be applied to traffic control signals in
which one or more, or even each phase has a fixed duration. While
the step of determining distribution data using the phase duration
data may not be necessary for a phase of fixed duration, this step
may still be performed, and the resultant data used together with
distribution data relating to a variable phase in obtaining the
probability data.
[0034] In embodiments in which a phase has a duration that is
variable in response to demand, the duration of the phase may be
variable in response to any measure indicative of demand. The
demand may based on traffic conditions. The demand is the actual
demand, e.g. resulting from actual traffic conditions. The demand
is not an expected or predicted demand. The demand is preferably a
current demand. The demand may be based upon traffic conditions
specifically at the intersection where the signal is located, or
may be based at least in part upon local traffic conditions. For
example, the traffic control signal may be one of a group of
traffic control signals which are together operated in a manner
that is responsive to demand.
[0035] The traffic control signal may be operated such that the
duration of the or each (variable) phase is variable in response to
sensed demand. The demand may be sensed by the traffic control
signal or by another entity. Another entity may then transmit data
indicative of the sensed demand to the traffic control signal or
data for controlling the signal in response to the sensed demand.
For example, in some preferred embodiments the duration of the or
each variable phase is variable in response to an amount of traffic
arriving at the signal. The traffic may be traffic that arrives at
the signal and passes through, or that arrives and is held in a
queue, depending upon the phase of the traffic signal concerned.
The amount of arriving traffic may be sensed using, for example,
fixed loop induction systems, a video surveillance system, etc.
Rather than being response to a sensed amount of traffic arriving
at the signal in this manner, variation of the duration of a phase
of the traffic signal in response to demand may be achieved in
other manners. For example, a traffic control signal may sense
arriving traffic by direct communication with an approaching
vehicle, e.g. with a navigation device of the vehicle. The traffic
control signal may be in bi-directional communication with a
vehicle so as to be able to transmit data indicative of the future
operation of the signal to the vehicle. In other arrangements the
traffic control signal may be managed remotely in response to
traffic conditions by a traffic management system, e.g. server.
Thus, in accordance with preferred embodiments of the invention,
the traffic signal may be any traffic signal that is operated in a
manner that the duration of at least one phase is responsive to
demand, whether the signal is arranged to autonomously operate in
this manner, e.g. through sensing of arriving traffic, or is caused
to operate in this manner under the control of remote management
system, which has access to traffic information. Any traffic
control signal of this type may be referred to as "dynamically
managed".
[0036] The data indicative of the durations of multiple instances
of the at least one phase of the signal may be directly or
indirectly indicative thereof. The data is preferably indicative of
the respective durations of each of a plurality of different
instances of the or each phase in a given time period, to enable
data indicative of a distribution of the durations of the or each
phase to be determined for at least a portion of the time period.
Thus, for each of the at least one phase considered, data
indicative of the durations of multiple instances of the phase,
e.g. in a given time period, is used to obtain the data indicative
of a distribution of the durations of the instances of the phase,
e.g. in at least a portion of the given time period. The number of
different instances of a given phase for which data is obtained
and/or the given time period may be selected as appropriate to
provide data of a desired level of reliability, taking into
account, e.g. the level of variability of phase durations for the
traffic signal involved.
[0037] The multiple instances of a given phase may be instances
associated with different, e.g. successive cycles of the signal.
The multiple instances of a phase are preferably successive
instances of a phase. The obtained data may indicative of the
duration of every instance of the or each phase in a given time
period. However, it will be appreciated that the duration of
certain instances of a phase might be disregarded, e.g. if the
phase has a length that exceeds a predetermined threshold. For
example, if traffic levels are low along the path controlled by a
particular signal that is one of a group of signals controlling
paths at an intersection, the signal may be "skipped", resulting in
an unusually long "stop" phase of the signal. Preferably duration
data is obtained for each one of a set of more than two instances
of each phase.
[0038] The data indicative of the durations of different instances
of a phase or phases of the traffic control signal is indicative,
for each of the at least one phase, of the durations of different
instances of the phase of the traffic control signal, preferably in
a given time period. The time period may be selected to have a
length appropriate to reflect a sufficient number of instances of
the or each phase that a meaningful distribution of the duration
data for each phase may be obtained, while avoiding the need to use
excessive amounts of data processing and/or storage power.
[0039] The time period may be a period in the recent past such that
the phase duration data is indicative of the relatively current
operation of the traffic signal, e.g. within the last five, ten,
fifteen or thirty minutes. The data may then be considered to be
"live" data. Live data may thus be thought of as data which is
relatively current and relates to the operation of the traffic
control signal within the last thirty, fifteen, ten or five
minutes. It is envisaged that data may be received and stored at
intervals to update corresponding previously stored data. For
example, the data may be updated every 5 minutes. While such
embodiments may allow more accurate predictions of the future
operation of the traffic control signal to be made, being based
upon the most recent operation of the signal, such techniques are
more demanding in terms of processing and storage power.
[0040] In other embodiments the time period may be a historical
time period. In this context the word "historical" should be
considered to indicate data that is not live, that is data that is
not directly reflective of the operation of the traffic control
signal at the present time or in the recent past (perhaps within
roughly the last five, ten, fifteen or thirty minutes).
[0041] The time period may be a time period corresponding to a
timeslot of interest. The timeslot may be a timeslot at a
particular time of day, day of the week and/or relating to a
particular expected traffic intensity, e.g. peak, off-peak, etc. It
is envisaged that duration data may obtained for a plurality of
different time periods, corresponding to different timeslots, e.g.
timeslots during the day, days of the week, traffic intensity
levels, etc. When implementing the method of the invention, the
duration data used should relate to a time period that will be
relevant to the period for which a prediction of the operation of
the traffic signal is to be made. For example, if a prediction is
required for a morning peak time, then a time period corresponding
to the morning peak may be used. The method may comprise obtaining
the duration data for the time period from stored duration data
relating to a plurality of different time periods. The given time
period may not be predefined. The time period may merely be the
time period which is defined between the earliest and latest phases
to which the duration data relates.
[0042] In some preferred embodiments the data indicative of the
durations of the multiple instances of a phase comprises data
indicative of a list of durations of each of the multiple instances
of the phase, e.g. in a given time period, and preferably of every
instance of the phase in the given time period. In other
embodiments the data may be otherwise indicative of durations of
instances, whether directly or indirectly. For example the obtained
data may comprise a list of transition times indicative of start
and end times for each instance of a phase, etc. If the traffic
signal operates in accordance with a predetermined cycle plan then
the data may be indicative of a cycle time and the timing of at
least one transition between phases for the signal.
[0043] The method may extend to the step of obtaining the data
indicative of the durations of multiple instances of the at least
one phase of the traffic control signal, e.g. in a given time
period. The duration data may be obtained from any source or
sources. The method may comprise receiving the data from a source
or sources. For example, the data may be obtained from a third
party data provider. The data may be obtained over any suitable
communications network, such as a vehicle-to-vehicle (V2V) and/or a
vehicle-to-infrastructure (V2I) communications network. In other
embodiments the data might alternatively or additionally be
received from a server. Alternatively or additionally the data may
be obtained based upon data transmitted by the or a traffic control
signal. Thus, in embodiments the data may be obtained from a third
party data provider, from a server, from a vehicle, or from the or
a traffic control signal. Alternatively or additionally, the data
may be obtained from positional data relating to the movement of
one or more device with respect to time along the path controlled
by the traffic control signal. Such data may be referred to as
"vehicle probe data". As described, for example, in WO 2013/060774
A1, the contents of which are incorporated herein by reference,
such data may be used to determine data indicative of one or more
times at which a transition between phases of a traffic control
signal occurred. Accordingly, such data may be used to obtain data
indicative of a duration of the phases of the traffic control
signal. In some embodiments the method may extend to the step of
generating the duration data, e.g. using "probe" data.
[0044] In embodiments it is envisaged that the duration data may be
received from a server for use in accordance with the invention.
The data may be received by a navigation device which then carries
out the steps involved in obtaining the distribution data and, in
accordance with the first and second aspects, the probability data.
The data may be received in response to a request by the navigation
device for the duration data relating to the traffic control
signal. The traffic control signal may be a signal along a route
being navigated. The route may be a pre-calculated route or an
expected route as determined by the navigation device. The signal
may be the next signal along the route being navigated. The server
may store duration data in respect of a plurality of traffic
control signals in a geographic region. In this way navigation
devices may then request data as needed when they need to obtain
probability data in respect of a particular traffic control signal.
It will be appreciated that the duration data stored by the server
in these embodiments may be received from any of the sources
described, e.g. from a third party, from a vehicle, from a traffic
control signal, from vehicle probe data, or combinations thereof.
Of course other arrangements are possible. For example, a server
might determine the distribution data using duration data stored by
the server, and optionally the probability data, and then provide
the data to a navigation device. The distribution data may be
determined in response to a request by a navigation device. In
other arrangements, all steps may be performed by a navigation
device. Any step or steps may be performed by a server, a
navigation device, or combinations thereof.
[0045] In accordance with the invention, the duration data is used
in determining, for the or each of the at least one phase, data
indicative of a distribution of the durations of the multiple
instances of the phase, preferably in at least a portion of a given
time period for which duration data is obtained. Where not
otherwise stated, the distribution data refers to the distribution
of durations of instances of a phase in at least a portion of a
given time period to which the duration data relates. Thus
distribution data is obtained for each of the at least one phase to
which the duration data relates. Where multiple phases are
considered, the distribution data determined for each phase may or
may not be distinct. For example, distribution data may be obtained
separately for each phase and then combined, e.g. summed or
integrated. Thus, references to obtaining distribution data, e.g.
of a particular form, for each of multiple phases, does not
necessarily imply that the distribution data is distinct. The
distribution data may be of any form. Preferably the data, for the
or each phase, is indicative of an empirical distribution of the
durations of the multiple instances of the phase (i.e. the or each
phase of the at least one phase considered). In other words, an
empirical distribution is obtained based on the duration data. The
method may comprise obtaining, for the or each phase, a
distribution function indicative of the distribution of the
durations of the multiple instances of the phase, e.g. in the given
time period, preferably an empirical distribution function. The
method may comprise obtaining, for the or each phase, data
indicative of a distribution profile of the durations of the
multiple incidences of the or each phase. While the distribution
data is preferably indicative of an empirical distribution, it is
envisaged that the step of determining the distribution data may
comprise data, for the or each phase, indicative of a model
distribution indicative of the distribution of the durations of
multiple instances of the phase based on the duration data for the
phase. The data may be indicative of a model distribution function
or profile. For example a model distribution function might be a
normal distribution function.
[0046] The distribution data may be determined based on the
duration data for the at least one phase in any suitable
manner.
[0047] As mentioned above, in preferred embodiments at least, the
data indicative of the durations of instances of the at least one
phase of the traffic control signal that is obtained relates to a
given time period, i.e. the durations of the incidences of the
phase in the time period. The distribution data that is obtained is
then indicative of the distribution of the durations of the or each
phase in at least a portion of the time period. In other words,
distribution data may be derived for only a portion of a time
period for which phase duration data is obtained. This may
facilitate processing of the data.
[0048] In accordance with the invention in its first and second
aspects at least, the distribution data is used in obtaining the
data indicative of a probability of the traffic control signal
having a given phase at one or more future time. It will be
appreciated that the distribution data may be determined as part of
a single step in which the probability data is obtained from the
duration data, and therefore may or may not form a distinct step to
the determining of the probability data. The or each time for which
the probability data is obtained is a future time. It will be
appreciated that the phase whose probability at the future time is
being determined, may or may not correspond to a phase for which
distribution data was obtained. For example, the distribution data
may be obtained for a green phase of a traffic light, and used to
predict the probability of the light having a red phase at a time
of interest. However, preferably the given phase to which the
probability data relates includes at least one phase for which
distribution data is determined. Preferably the phase is a phase
which allows vehicles to pass along a path controlled by the
traffic signal, i.e. a "go" phase, such as a green light. The
probability data is then indicative of the probability of the
signal having a phase allowing traffic to pass, e.g. a "go" or
green phase.
[0049] The probability data is based upon the distribution data
relating to the or each phase for which such data is determined.
Thus, where the distribution data is obtained for multiple phases,
the distribution data relating to each phase is used in obtaining
the probability data.
[0050] The distribution data may be used alone in determining the
probability data, or together with other data. In some preferred
embodiments the method comprises using indicative of a timing of at
least one instance of a phase of the traffic control signal, e.g.
in a given time period to which the duration data relates together
with the duration data in obtaining the probability data. The
method may extend to obtaining the timing data. The timing data is
preferably indicative of a transition time associated with the
instance of the phase, e.g. a time of a transition to or from the
phase from or to another phase. For example, the timing data may be
indicative of a time at which a transition to a current phase of
the traffic signal occurred. Timing data need only be provided in
respect of one of the phases for which distribution data is
determined where such data is obtained for multiple phases.
However, timing data may be provided for each phase, or at least
for multiple instances of one or more, or each phase. By providing
timing data, a reference point for the duration data for the at
least one phase with respect to time is provided. This may
facilitate obtaining the probability data. However, it will be
appreciated that specific timing data associated with the duration
data need not necessarily be obtained, as, some timing data will be
inherent in the duration data. In some embodiments the probability
data is obtained using the obtained duration data and data
indicative of a current state of the signal, the current state data
being indicative of a current phase of the signal and a time of
transition to the current phase.
[0051] The probability data may be obtained using the distribution
data in any suitable manner. It will be appreciated that obtaining
the probability data involves converting data indicative of a
distribution of durations for a phase or phases, e.g. in a given
time period to data indicative of the expected probability of the
signal having a phase at one or more future time. It will be
appreciated that the probability data for a given phase may be
obtained using distribution data relating to the phase, or another
phase, or a combination thereof. However, preferably at least
distribution data relating to the phase for which probability data
is to be determined is used.
[0052] In some preferred embodiments the method comprises using the
distribution data to obtain data indicative of the probability of
the traffic control signal having the given phase with respect to
time, i.e. future time. In embodiments the method may comprise
obtaining data indicative of the probability of the traffic control
signal having the given phase with respect to time over a given
future time period. Thus, the obtained probability data may be
indicative of the probability of the traffic control signal having
a given phase with respect to time over a given future time period.
However, it will be appreciated that probability data may instead
be determined individually for specific time(s) of interest. The
probability data is preferably indicative of the probability of the
signal having the given phase over a continuous range of time. In
embodiments, the probability data is indicative of the variation in
probability with respect to time.
[0053] The probability data may be indicative of a probability
function. For example, the probability function may be a
probability mass or probability density function indicative of the
probability of the signal having the given phase with respect to
time. The probability function is preferably indicative of the
probability of the signal having the given phase over a continuous
range of time. For example, the probability data may be in the form
of a plot e.g. curve indicating probability with respect to
time.
[0054] The probability data may be obtained by determining the
probability that each of said one or more future time falls within
said given phase of said traffic control signal.
[0055] The step of determining the probability data for the or each
future time may comprise determining the probability of the or each
future time coinciding with the given phase for each of a plurality
of possible cycle plans of the traffic control signal. In
embodiments the probability data may be obtained by combining, e.g.
summing the respective probabilities of the time coinciding with
the given phase in each of the plurality of possible cycle plans.
That is, to determine the probability of a given phase at one or
more times in the future, the probability of said one or more times
falling within the given phase in the first possible cycle plan,
second possible cycle plan, and so on may be summed. The
probability of each possible cycle plan having a given phase at one
or more times in the future can be calculated based on the
distribution data. A "possible" cycle plan refers to a cycle plan,
i.e. cycle length and phases within the cycle, which may fit the
duration data.
[0056] In embodiments in which data indicative of the probability
of the signal having the given phase with respect to time is
determined, the method may comprise identifying data indicative of
one or more turning points in the probability with respect to time,
and determining a time associated with the or each turning point,
e.g. with the position thereof. The or each turning point may be a
maximum or minimum. Preferably data indicative of one or more
maxima is identified and the corresponding time(s) determined. In
these embodiments the probability data is preferably indicative of
probability over a continuous range of time. Thus, the method may
comprise obtaining data indicative of the probability of the
traffic control signal having the given phase with respect to time
over a given time period, wherein the probability data is
indicative of one or more, and preferably a plurality of turning
points, e.g. maxima or minima, with respect to time. A maximum or
minimum in the probability will correspond to times at which it is
determined to be most or least likely that the signal will have the
given phase respectively. It will be appreciated that there may be
multiple maxima and minima associated with times at which the
probability has maximum and minimum values respectively over
successive cycles. While the turning points may be turning points
in a plot of probability against time, it will be appreciated that
such a plot may not necessarily be derived, and the turning points,
and their associated times, may simply be determined directly from
the probability against time data. In some embodiments the time
associated with a turning point is associated with a position of
the turning point, i.e. in a plot of probability with respect to
time.
[0057] The obtained probability data may be used in various
manners. In some preferred embodiments the method comprises using
the determined probability data to provide a speed recommendation
for a vehicle. The vehicle may be a vehicle approaching the traffic
signal. The method may comprise using the probability data to
provide a speed recommendation for a vehicle to enable the vehicle
to arrive at the traffic control signal at or around a time which
is expected to coincide with a phase of the signal allowing the
passage of traffic along a path controlled by the traffic control
signal based on the probability data. The time expected to coincide
with a phase of the signal allowing the passage of traffic is
preferably a time at which the control signal is most likely to
have a phase allowing the passage of traffic based on the
probability data. In preferred embodiments as described above, the
method comprises determining data indicative of the probability of
the signal having the given phase with respect to time, and
determining a time or times associated with one or more turning
points, preferably maxima, identified in the probability. The
time(s) may be associated with the position(s) of the or each
turning point. The probability data is preferably indicative of the
probability of the signal having a phase allowing the passage of
traffic, i.e. a "go" or "green" phase. The time associated with the
maxima will then be indicative of the most likely times of phases
allowing the passage of traffic. The time may be identified by
consideration of a position of the maximum in a plot of probability
with respect to time. The method may comprise using the data
indicative of one or more turning points, e.g. maxima, in providing
the speed recommendation. The method may comprise using one or more
times associated with identified turning points, e.g. maxima, in
the probability data, in providing the speed recommendation. The
one or more times may be associated with positions of the one or
more maxima. The method may comprise providing a speed
recommendation for a vehicle which will enable the vehicle to
arrive at the traffic control signal at or around a time which
coincides with a time associated with an identified turning point
e.g. maximum in the probability data. Where the probability data
comprises a plurality of maxima having associated with different
respective times, the method may comprise selecting one of the
times upon which the speed recommendation is to be based, to enable
the vehicle to arrive at the traffic control signal at that time.
The selection of the time may be based upon a current speed of the
vehicle, and/or a speed limit governing the route from a current
location to the traffic control signal. For example, the selected
time may be one that may be achieved with minimal adjustment of the
speed of the approaching vehicle, and/or which allows the vehicle
to travel at a speed close to the relevant speed limit for the
approach.
[0058] It is believed that using data indicative of a distribution
of multiple instances of at least one phase of a traffic control
signal in providing a speed recommendation is advantageous in its
own right, whether or not that data is first used to determine
probability data as described above.
[0059] In accordance with a further aspect of the invention there
is provided a method of providing a speed recommendation for a
vehicle to enable the vehicle to arrive at a traffic control signal
at a time expected to coincide with a phase allowing the passage of
traffic along a path controlled by the traffic control signal,
wherein the speed recommendation is obtained using data indicative
of a distribution of the durations of multiple instances of at
least one phase of the traffic control signal.
[0060] In these embodiments the at least one phase of the traffic
control signal for which distribution data is obtained preferably
includes a phase allowing the passage of traffic. The traffic
control signal is operable to transition between phases in use,
which phases include a phase that allows the passage of traffic
along a path controlled by the traffic signal, and a phase that
prevents the passage of traffic along a path controlled by the
traffic control signal. The method may extend to the step of
determining the distribution data using data indicative of the
durations of multiple instances of the at least one phase of the
traffic control signal, e.g. in a given time period.
[0061] In accordance with a further aspect of the invention there
is provided a system for providing a speed recommendation for a
vehicle, the system comprising:
[0062] means for providing a speed recommendation for a vehicle to
enable the vehicle to arrive at a traffic control signal at a time
expected to coincide with a phase allowing the passage of traffic
along a path controlled by the traffic control signal, wherein the
speed recommendation is obtained using data indicative of a
distribution of the durations of multiple instances of at least one
phase of the traffic control signal.
[0063] The present invention in these further aspects may include
any or all of the features described in relation to the first and
second aspect of the invention, and vice versa, to the extent that
they are not mutually inconsistent. Thus, if not explicitly stated
herein, the system of the present invention may comprise means for
carrying out any of the steps of the method described. In these
further aspects, any of the steps and features, including, for
example, the duration data, or the steps involved in obtaining the
distribution data may be in accordance with any of the embodiments
described above in relation to the first and second aspects of the
invention. Thus, any of the features described by reference to
these further aspects of the invention may be used in the earlier
aspects of the invention, and vice versa, to the extent they are
not mutually inconsistent.
[0064] The system and method in these further aspects may be
implemented by a server or a navigation device, or combinations
thereof. For example a navigation device may determine the speed
recommendation using distribution data obtained from a server, or
may additionally determine the distribution data using duration
data, which may be received from a server. Alternatively a speed
recommendation may be determined by a server and transmitted to a
navigation device.
[0065] In accordance with any of the aspects or embodiments of the
invention involving determining a speed recommendation, including
those described below based on expected waiting time, other data
may be used in determining the speed recommendation. For example a
current position of the vehicle may be used, and optionally data
indicative of a current speed of the vehicle. The method may
comprise additionally using data indicative of a timing of at least
one instance of a phase of the traffic control signal, e.g. in a
given time period, to which the duration data relates together with
the duration data in obtaining the speed recommendation. The method
may extend to obtaining the timing data. The timing data is
preferably indicative of a transition time associated with the
instance of the phase, e.g. a time of a transition to or from the
phase from or to another phase. For example, the timing data may be
indicative of a time at which a transition to a current phase of
the traffic signal occurred. Timing data need only be provided in
respect of one of the phases for which distribution data is
determined where such data is obtained for multiple phases. In some
embodiments the speed recommendation is obtained using the obtained
duration data and data indicative of a current state of the signal,
the current state data being indicative of a current phase of the
signal and a time of transition to the current phase.
[0066] In accordance with any of the aspects or embodiments of the
invention involving determining a speed recommendation, including
those below based upon expected waiting time, the speed
recommendation may be of any suitable form. Method of providing
speed recommendations to vehicles are described, for example, in WO
2012/034582 A1 entitled "Improvements in or relating to portable
processing devices"; the contents of which are herein incorporated
by reference. The speed recommendation may be a recommendation of a
single target speed. However, preferably the speed recommendation
is in the form of a recommendation of a range of speed. The range
of speed is a range within which it is determined a driver may
travel in order to arrive at the traffic control signal to coincide
with the given phase, e.g. a phase allowing the passage of traffic.
The method may comprise providing the speed recommendation as a
recommended speed window. The range of speed may be a range of
speed which will result in the vehicle arriving at the control
signal within given range around a time at which the signal is most
likely to have a phase allowing the passage of traffic based on the
probability data, e.g. a time associated with the position of a
maxima. The time range around a most likely time may be a fixed
time range, e.g. as a fraction of a cycle length, or may be
determined by reference to the probability, e.g. being a time range
in which the probability of the phase being one allowing the
passage of traffic is above a given level.
[0067] The method may comprise outputting the speed recommendation
to a driver or an Advanced Driver Assistance System (ADAS). This
may be carried out in any suitable manner. The method may comprise
displaying the speed recommendation to a driver. The speed
recommendation may be displayed by displaying a recommended speed
window. The method may comprise providing a graphical indication of
the recommended speed to the driver. The step of providing a speed
recommendation to a driver or ADAS may be carried out by a
navigation device.
[0068] Alternatively or additionally, in some embodiments the
method of the first and second aspects may comprise using the
obtained probability data indicative of the probability of the
traffic control signal having a given phase at a future time or
times to determine an expected waiting time for a vehicle when
arriving at the signal at one or more future time of interest. The
future time of interest will be a time for which probability data
has been determined, whether a discrete future time or a time in a
range of time for which probability data has been determined. The
expected waiting time is indicative of the time that a vehicle is
expected to have to wait at the signal for a phase allowing the
vehicle to pass along the path controlled by the signal when
arriving at the time.
[0069] The expected waiting time may be obtained using the
probability data in any suitable manner. It will be appreciated
that such a time may be determined using probability data that is
in relation to the probability of the signal having a given phase
that is a "stop" phase or a "go" phase, as each will be directly or
indirectly indicative of the timing of a "go" phase, i.e. where the
probability relates to the "stop" phase, lower probabilities will
be indicative of a greater likelihood of a "go" phase being
encountered at the time. Preferably, however, the probability data
is in relation to a phase allowing the passage of traffic, i.e. a
"go" phase, such as a green light. In embodiments in which the
method comprises identifying data indicative of one or more turning
points, e.g. maximum or minimum, in the probability data with
respect to time, the method may comprise using a time associated
with an identified turning point, e.g. maximum or minimum, in
determining the expected waiting time. Preferably a maximum is
identified and used. The probability data then relates to a phase
allowing the passage of vehicles along a path controlled by the
traffic control signal. For example, the expected waiting time
might be based upon a time difference between the time of interest
for which the waiting time is being determined and the time at
which a maximum in the probability occurs, e.g. indicative of the
next likely "go" phase. Determination of expected waiting time may
be carried out in other manners, and may involve more complex
analysis. The expected waiting time may take into account the
differing amounts of time that a vehicle might have to wait for a
phase allowing it to pass under the various possible phase
scenarios that may be encountered at that time. The data may
involve using the probability data obtained in respect of one or
more phases of the traffic control signal with respect to time.
[0070] It is believed that determining expected waiting time data
based on data relating to a distribution of the durations of
multiple instances of a phase of a traffic signal is advantageous
in its own right.
[0071] In accordance with a further aspect of the invention there
is provided a method for determining data indicative of an expected
waiting time for a vehicle arriving at a traffic control signal,
the traffic control signal being operable to transition between
different phases in use, the method comprising:
[0072] obtaining the data indicative of an expected waiting time
for a vehicle arriving at the traffic signal at one or more future
times using data indicative of a distribution of the durations of
multiple instances of at least one phase of the traffic control
signal.
[0073] In accordance with a further aspect of the invention there
is provided a system for determining data indicative of an expected
waiting time for a vehicle arriving at a traffic control signal,
the traffic control signal being operable to transition between
different phases in use, the system comprising:
[0074] means for obtaining the data indicative of an expected
waiting time for a vehicle arriving at the traffic signal at one or
more future times using data indicative of a distribution of the
durations of multiple instances of at least one phase of the
traffic control signal.
[0075] The present invention in these further aspects may include
any or all of the features described in relation to the first or
second aspects of the invention, and vice versa, to the extent that
they are not mutually inconsistent. Thus, if not explicitly stated
herein, the system of the present invention may comprise means for
carrying out any of the steps of the method described. In these
further aspects, any of the steps and features, including, e.g. the
duration data, the steps involved in obtaining the distribution
data may be in accordance with any of the embodiments described
above in relation to the first and second aspects of the invention.
Thus, any of the features described by reference to these further
aspects of the invention may be used in the earlier aspects of the
invention, and vice versa, to the extent they are not mutually
inconsistent.
[0076] It will be appreciated that the duration data indicative of
the durations of different instances of a particular phase or
phases of the traffic light may enable an expected waiting time to
be determined, based upon the possible phase scenario that may be
encountered when arriving at the signal at a future time or
times.
[0077] In these embodiments the at least one phase of the traffic
control signal for which distribution data is obtained preferably
includes a phase allowing the passage of traffic. The traffic
control signal is operable to transition between phases in use,
which phases include a phase that allows the passage of traffic
along a path controlled by the traffic signal, and a phase that
prevents the passage of traffic along a path controlled by the
traffic control signal. The method may extend to the step of
determining the distribution data using data indicative of the
durations of multiple instances of the at least one phase of the
traffic control signal.
[0078] The system and method in these further aspects may be
implemented by a server or a navigation device, or combinations
thereof. For example a navigation device may determine the expected
waiting time data using distribution data obtained from a server,
or may additionally determine the distribution data using duration
data, which may be received from a server. Any other arrangement
may be used, however.
[0079] The method may comprise additionally using data indicative
of a timing of at least one instance of a phase of the traffic
control signal, e.g. in a given time period, to which the duration
data relates together with the duration data in obtaining the
expected waiting time data. The method may extend to obtaining the
timing data. The timing data is preferably indicative of a
transition time associated with the instance of the phase, e.g. a
time of a transition to or from the phase from or to another phase.
For example, the timing data may be indicative of a time at which a
transition to a current phase of the traffic signal occurred.
Timing data need only be provided in respect of one of the phases
for which distribution data is determined where such data is
obtained for multiple phases. In some embodiments the expected
waiting time data is obtained using the obtained duration data and
data indicative of a current state of the signal, the current state
data being indicative of a current phase of the signal and a time
of transition to the current phase.
[0080] In accordance with any of the aspects or embodiments of the
invention in which expected waiting time at a given traffic control
signal is determined, the time of interest for which expected
waiting time is obtained, i.e. the time of arrival at the signal,
may be a time at which the vehicle is expected to arrive at the
traffic control signal when following a given route. The route is a
route which involves passing along the path controlled by the
traffic control signal, and may be any route from a first location
to a second location. The route may be a pre-calculated route to a
destination, or a portion thereof, or a route that it is expected
the vehicle will follow, e.g. based upon a current trajectory,
previous history, etc. The route may be a route being navigated, or
yet to be navigated. The route might be an alternative route to at
least a portion of an existing route being navigated. Thus the
first location may be a current position, a position ahead of a
current position along a route being navigated, or an origin. The
second location may be a destination, or a position ahead of a
current position along a route being navigated. The first and
second locations may be automatically determined, or may be user
specified, or combinations thereof.
[0081] The expected waiting time may then be used in obtaining a
more accurate estimate as to expected travel time along the route
e.g. from the first location to the second location. The method may
further comprise using the determined expected waiting time in
determining an estimated travel time for the route. The expected
waiting time is indicative of the time delay that may be expected
at the signal. This provides the ability to more accurately
predicted travel times along routes and/or to estimate times of
arrival than prior art techniques which could not take into account
the likely phase of signals along the route, and would merely add
arbitrary delay times to estimated route timings to account for the
presence of signals along the route. The present invention accounts
for the fact that, depending upon arrival time, a given signal may
or may not give rise to a delay.
[0082] In some embodiments the method may further comprise
determining an expected queue time at the traffic control signal,
the expected queue time being indicative of the time a vehicle can
be expected to queue before passing through the signal when
arriving at the signal at an expected time. If there is a high
traffic intensity, then a vehicle may have to queue before passing
through the signal even if they arrive so as to coincide with a
phase allowing the passage of vehicles. The method may comprise
adjusting an expected waiting time at a signal and/or travel time
for a route including a signal to account for expected queue time
at the signal. Any of the methods described herein utilising
expected waiting time at a signal or signals may also take into
account expected queue time when arriving at the signal. Thus an
additional delay factor may be taken into account. Queue times may
be determined using e.g. historical vehicle probe data, and may be
derived for different time periods in the day.
[0083] In accordance with the invention in any of its aspects and
embodiments, the method may comprise carrying out any of the steps
involved in obtaining an expected waiting time for a given traffic
control signal in accordance with any of the aspects of the
invention, in respect of a plurality of traffic control signals
which control traffic flow along a path included in a route. The
route may be of any of the types outlined above. The method may
then comprise using each determined expected waiting time in
determining an estimated travel time for the route. In embodiments
in which an expected waiting time for one or more further traffic
control signal is used, the waiting time at the or each subsequent
traffic control signal may be dependent upon the expected waiting
time determined for the or each previous traffic control
signal.
[0084] In any of the embodiments in which expected waiting time is
determined for one or more time of interest, the method may
comprise obtaining data indicative of the expected waiting time
with respect to time of arrival at the signal at a plurality of
times in a given future time period, e.g. over a given future time
period. The data indicative of the expected waiting time with
respect to time is indicative of a variation in expected waiting
time with respect to time of arrival when arriving at the signal at
different times in the future time period. The data is preferably
indicative of the expected waiting time over a continuous future
time period. Such data may be obtained based on probability data
with respect to time, when determined, or directly from
distribution data. The data indicative of estimated waiting time
with respect to time may be indicative of one or more turning
points, e.g. maxima and minima. This reflects the cyclic nature of
the operation of the signal.
[0085] In accordance with the preferred embodiments in which data
indicative of expected waiting time with respect to time of arrival
at a signal at different times over a given future time period is
obtained, the method preferably comprises using the data to
determine a speed recommendation for a vehicle. The speed
recommendation is preferably one which is expected to minimise
expected waiting time when arriving at the signal. The speed
recommendation will result in the vehicle arriving at the signal at
a time that minimises expected waiting time. The waiting time is
"minimised" as determined using the expected waiting time data.
[0086] In some preferred embodiments data indicative of an expected
waiting time with respect to time of arrival at the signal is
obtained for each of a plurality of traffic control signals which
control traffic flow along a path included in a route. The route
may be of any of the types described above. The method may then
comprise obtaining a speed recommendation for a vehicle travelling
along the route which will minimise expected waiting time when
travelling along the route based on the expected waiting time data
obtained for each traffic control signal.
[0087] The steps of determining a speed recommendation to minimise
expected waiting time along a route, or using expected waiting time
to obtain a travel time along a route, may be carried out by a
navigation device. The navigation device may carry out such steps
using expected waiting time data provided by a server, or may
itself determine the expected waiting time data. In the latter case
the device may determine the expected waiting time using duration
data provided by the server. For example, the method may comprise
the navigation device providing data indicative of a route being
followed to a server, and the server then providing the necessary
data relating to traffic control signals along the route to the
navigation device for use by the device. This may minimise the
amount of data the navigation device need store, as the relevant
data may be obtained from a server for those traffic control
signals of interest. However, other arrangements are possible.
[0088] In accordance with yet further embodiments data indicative
of an expected waiting time is obtained in respect of one or more,
and preferably a plurality of traffic control signals associated
with navigable segments of a navigable network, and used in
generating a route through the navigable network. The route is a
route from a first location to a second location and may be of any
of the types discussed above. The second location may be a
destination. The first location may be a current location or
origin. The route is preferably a fastest route. In this way, the
expected waiting time, indicative of a delay which may be incurred
a traffic control signals which may be included in a route through
the network is taken into account when generating a fastest route.
The method may comprise using the expected waiting time data to
generate a route through the navigable network which minimises
expected waiting time at traffic control signals along the route.
Again, such steps may be carried out by a server or a navigation
device. In some embodiments the method may comprise a server
generating the route and data indicative of the route for
transmission to a navigation device. The server may transmit the
data. The route may be generated at the request of the navigation
device. For example, the device may provide the server with data
indicative of the first and second locations for the route, with
the server then generating a route by reference to expected waiting
time. The data indicative of the route may be for transmission in
any manner, and may or may not be directly indicative of the route
itself. A hybrid route generation process between the server and
navigation device may be used. For example, data may be generated
for transmission that will cause the relevant navigable segments
for inclusion in the route to be favoured when a routing engine of
the navigation device itself generates a route. Of course, a
navigation device might alternatively generate a route without
interaction with the server.
[0089] Any references to a route herein may refer to a
pre-calculated route or portion thereof, or any other path being
followed or to be followed, e.g. an estimated route, etc. The route
may be a route to a destination. The route may be a route being
navigated or may be a planned route. The route is from a first
location to a second location. The first location may be a current
position, a position at or ahead of a current position along a
route being navigated, or an origin of a route. The second location
may be a destination, or may be a location ahead of a current
position along a route being navigated.
[0090] As mentioned above, the steps of the method in accordance
with the invention in its various aspects and embodiments may be
carried out by different devices and/or in different locations. In
preferred embodiments the methods are carried out by a server or a
navigation device, or may be carried out in part by a navigation
device and in part by a server. The way in which the steps of the
methods are split between a server and navigation device may be
selected as desired, and may be chosen to provide a balance as to
demands upon data processing and/or storage power, and reducing
bandwidth. While using a server to carry out certain of the
processing steps may reduce the data processing and/or storage
demands placed on a navigation device, this will necessitate the
transmission of greater quantities of data between the server and
navigation device, which may use bandwidth.
[0091] In some embodiments the method may comprise a navigation
device receiving the data indicative of the durations of multiple
instances of the at least one phase of the traffic control signal
in a given period, which is used in the invention in its various
aspects and embodiments, from a server. The method may comprise a
server transmitting such data to a navigation device. In some
embodiments the method is carried out by a navigation device, and
further comprises the step of the navigation device obtaining the
duration data from a server. The method may comprise a server
storing the duration data. The server may store data indicative of
the durations of multiple instances of at least one phase of the
traffic control signal, for a plurality of traffic control signals,
e.g. in a given geographic area.
[0092] The step of using the duration data to obtain distribution
data in accordance with the invention in any of its aspects or
embodiments may be carried out by a server or a navigation device,
but in some preferred embodiments is carried out by a navigation
device. Preferably the step is carried out using duration data
obtained from a server. It has been found that this may be more
efficient in terms of processing power, as the navigation device
need not then store the duration data for the traffic control
signal. The navigation device may request the duration data for a
given traffic control signal from the server. The traffic control
signal may be the next traffic control signal along a route being
navigated, whether a pre-calculated route or not. The navigation
device may then carry out the steps of determining probability
data, and/or an expected waiting time, speed recommendation or
route recommendation using the distribution data as appropriate in
the various aspects of the invention. In other embodiments,
determining the probability data might also be carried out by a
server, with a navigation device then simply carrying out the steps
involved in using that data, e.g. obtaining a speed recommendation,
expected waiting time, route recommendation, etc.
[0093] The methods of the present invention in any of their aspects
or embodiments may be implemented in relation to one or more
traffic control signals. The traffic control signal may be any
traffic control signal whose operation is of interest. In some
embodiments the or each traffic control signal is a traffic control
signal which controls traffic flow along a path included in a
route. In accordance with the invention in any of its aspects and
embodiments, the method may comprise carrying out the steps of
obtaining probability data, and/or expected waiting time data, in
respect of a plurality of traffic control signals which control
traffic flow along a path included in the route.
[0094] In these further aspects and embodiments of the invention in
which an expected waiting time is determined, the method may
comprise obtaining data indicative of the expected waiting time for
a vehicle when arriving at the traffic control signal at one or
more times of interest. Data indicative of expected waiting time
may be determined with respect to time over a given time period.
For example, the corresponding expected waiting time may be derived
for the or each time for which probability data is determined in
embodiments where it is based upon the probability data.
[0095] In any of the embodiments in which a route is generated, a
speed recommendation determined, a travel time determined, etc, for
example using expected waiting time data, the method may comprise
outputting and/or storing the determined route, recommendation or
travel time. The output may be to a driver or an ADAS. The step of
outputting may be in accordance with any of the embodiments
described for outputting a speed recommendation above. The method
may comprise displaying the route, recommendation or time to a
driver. The outputting step may be carried out by a navigation
device. The method may comprise providing a set of navigation
instructions for guiding a driver along a generated route.
[0096] In certain embodiments discussed above, the duration data
may be based upon data indicative of the position of a plurality of
devices with respect to time, i.e. probe data. The positional data
used in accordance with these embodiments of the invention may be
collected from one or more, and preferably multiple devices, and
relates to the movement of the devices with respect to time. Thus,
the devices are mobile devices. It will be appreciated that at
least some of the positional data is associated with temporal data,
e.g. a timestamp. For the purposes of the present invention,
however, it is not necessary that all positional data is associated
with temporal data, provided that it may be used to provide the
information relating to the duration of phases of a traffic control
signal. However, in preferred embodiments all positional data is
associated with temporal data, e.g. a timestamp.
[0097] The positional data relates to the movement of the or each
device with respect to time, and may be used to provide a
positional "trace" of the path taken by the device. The data may be
received from the device(s) or may first be stored. The devices may
be any mobile devices that are capable of providing the positional
data and sufficient associated timing data for the purposes of the
present invention. The device may be any device having position
determining capability. For example, the device may comprise means
for accessing and receiving information from WiFi access points or
cellular communication networks, such as a GSM device, and using
this information to determine its location. In preferred
embodiments, however, the device comprises a global navigation
satellite systems (GNSS) receiver, such as a GPS receiver, for
receiving satellite signals indication the position of the receiver
at a particular point in time, and which preferably receives
updated position information at regular intervals. Such devices may
include navigation devices, mobile telecommunications devices with
positioning capability, position sensors, etc. The device may be
associated with a vehicle. In these embodiments the position of the
device will correspond to the position of the vehicle. The device
may be integrated with the vehicle, or may be a separate device
associated with the vehicle such as a portable navigation
apparatus. Of course, the positional data may be obtained from a
combination of different devices, or a single type of device.
[0098] The positional data obtained from the plurality of devices
is commonly known as "probe data". The data obtained from devices
associated with vehicles may be referred to as vehicle probe data.
References to "probe data" herein should therefore be understood as
being interchangeable with the term "positional data", and the
positional data may be referred to as probe data for brevity
herein.
[0099] A navigation device as referred to herein may be a vehicle
based navigation device, and may be a PND or integrated device.
[0100] It will be appreciated that the methods in accordance with
the present invention may be implemented at least partially using
software. It will this be seen that, when viewed from further
aspects, the present invention extends to a computer program
product comprising computer readable instructions adapted to carry
out any or all of the method described herein when executed on
suitable data processing means. The invention also extends to a
computer software carrier comprising such software. Such a software
carrier could be a physical (or non-transitory) storage medium or
could be a signal such as an electronic signal over wires, an
optical signal or a radio signal such as to a satellite or the
like.
[0101] The present invention in accordance with any of its further
aspects or embodiments may include any of the features described in
reference to other aspects or embodiments of the invention to the
extent it is not mutually inconsistent therewith.
[0102] Advantages of these embodiments are set out hereafter, and
further details and features of each of these embodiments are
defined in the accompanying dependent claims and elsewhere in the
following detailed description.
BRIEF DESCRIPTION OF THE DRAWINGS
[0103] Various embodiments of the present invention will now be
described, by way of example only, and with reference to the
accompanying drawings in which:
[0104] FIG. 1 shows a flow diagram of a method in accordance with a
preferred embodiment of the present invention;
[0105] FIG. 2A shows measured green and red times in the period
from 15:00 till 19:00 for a particular signal group at an
intersection;
[0106] FIG. 2B shows histograms of the green and red times for the
same intersection during evening rush hour, where at least the red
times are not normally distributed;
[0107] FIG. 2C shows histograms of the green and red times for the
same intersection from 9:30 till 14:30, during which the times are
approximately normally distributed;
[0108] FIG. 3 shows the probability of green signal as a function
of time, and the conversion into a visual prediction;
[0109] FIG. 4 shows how the converted predictions of FIG. 3 for a
number of sequentially traversed traffic control signals can be
used to provide speed recommendations to a driver;
[0110] FIGS. 5A to 5C show the expected waiting times at three
intersections;
[0111] FIG. 6 illustrates how the expected travel time i.e. sojourn
time can be determined from the expected waiting times at
sequential intersections;
[0112] FIG. 7 shows the position of the intersections shown in
FIGS. 5A to 5C;
[0113] FIG. 8 shows an expected sojourn time through the network
shown in FIG. 7;
[0114] FIG. 9 shows the expected sojourn time for the route shown
in FIG. 7 for all possible arrival times in the range
considered;
[0115] FIG. 10 shows how the processing can be distributed between
a navigation device and a server; and
[0116] FIGS. 11A and 11B illustrate the effect of low and heavy
traffic flows on queue time at a traffic control signal.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
[0117] Some preferred embodiments of the invention will now be
described. The present invention will be described by reference to
a traffic control signal of a traffic control system. The traffic
control signal may be, for example, a traffic signal of a traffic
signal controlled crossing. The traffic control signal controls the
flow along a particular path e.g. at an intersection. The traffic
control signal may form part of a group of traffic controls signals
at the intersection. The embodiments of the invention will be
described, by way of example, with reference to a traffic control
signal that is responsive to vehicle demand. Such a traffic control
signal may be referred to as "dynamically managed". The traffic
control signal has red and green phases whose durations are
variable within predefined limits in response to vehicle demand.
For example, where there is a high traffic intensity along a road
segment approaching the traffic control signal, the duration of a
green phase may be increased to allow the passage of a greater
number of vehicles. The traffic control signal may be arranged to
sense approaching vehicles in some manner. For example, the traffic
control signal may be equipped with a suitable vehicle detection
means, e.g. a camera. In other arrangements the signal may be able
to communicate wirelessly with approaching vehicles which are
equipped with a suitable communications module. The communication
may be bi-directional, so as to allow the traffic control signal to
impart information regarding its operation to the vehicles, e.g.
upcoming green times, etc. It will be appreciated that the traffic
control signal may also have a yellow phase. For the purposes of
the present invention, the yellow phase is considered as part of
the red or the green phase, depending upon whether it allows the
passage of vehicles according to the relevant local law.
[0118] Although the present invention is described in conjunction
with a traffic control signal whose phase durations are variable in
response to demand, it will be appreciated that the techniques of
the present invention may equally be applied to traffic control
signal having only one phase, e.g. a green phase, whose duration is
variable in response to demand, or even one which has no such
phases, with each phase being fixed in accordance with a
predetermined cycle plan (with the durations of the durations of
the phases according to the predetermined cycle plan potentially
then being time dependent, e.g. to provide different durations for
a given phase in off peak times by comparison to peak times). Such
traffic control signals may be referred to as "statically
managed".
[0119] An embodiment of the invention which allows information
about the operation of a traffic control signal to be obtained will
now be described by reference to FIG. 1. More specifically, this
embodiment is used to predict the likelihood of the traffic control
signal having a particular phase, e.g. a green phase, at a future
time.
[0120] In accordance with step 1, lists indicative of the durations
of different instances of each of the red and green phases of the
traffic control signal in a given time period are obtained. These
may be obtained in various manners. In some embodiments the lists
of green and red times may be obtained using sources of data which
may include any one or ones of: third party data, vehicle probe
data, vehicle-to-vehicle (V2V) and/or vehicle-to-infrastructure
(V2I) data, and data obtained from the traffic control signal
itself. Vehicle probe data is vehicle probe data received from
devices associated with vehicles, e.g. equipped with navigation
satellite systems, such as GPS devices, whose position corresponds
to that of the vehicle. The probe data may alternatively be
referred to as "positional data". The probe data is associated with
temporal data, and may be used, for example, to derive probe traces
relating to the travel of probe vehicles in a geographic region
that includes the traffic control signal of interest. Such data may
be used to infer the points in time at which the traffic control
signal had a green phase or a red phase, e.g. as described in WO
2013/070774 A1. The duration lists may be derived by a server,
which may also be arranged to collect the relevant data, e.g.
through communication with traffic control signals, devices,
vehicles, etc.
[0121] The lists of durations may include the duration of each
instance of each of the red and green phases in the relevant time
period. However, the duration of certain phase instances may be
omitted if they are determined not to correspond to phase durations
for the normal operation of the traffic control signal. For
example, if the traffic intensity is low along a particular path
controlled by a traffic control signal which has phase durations
variable in response to demand, a cycle of the traffic control
signal may be omitted to allow greater numbers of vehicles to pass
along paths controlled by other traffic control signals forming
part of a signal group including that traffic control signal. To
avoid such situations distorting the duration data, red phase
durations which are significantly greater than the cycle length may
be omitted.
[0122] The duration lists for the red and green phases include the
durations of different instances of each of the phases in a given
time period. The time period may be a time period of interest, i.e.
corresponding to the approximate timeslot to which the desired
prediction is to relate. Duration lists may be stored, e.g. by a
server, and, where appropriate derived, for each of a plurality of
different time periods. For example, the time periods may include
any or all of: night, morning rush hour, off peak
morning/afternoon, evening rush hour and off peak evening. In other
arrangements, red and green duration lists may be stored in
relation to a time period corresponding to the previous 30 minutes
and updated every 5 minutes. This more dynamic arrangement may
provide better accuracy in predicting the operation of the traffic
control signal, reflecting the most recent operation of the traffic
control signal. However it would be more demanding in terms of
processing power. The time period may be selected to have a length
that is appropriate to provide a balance between accuracy of
prediction, and use of processing and/or storage power. Where a
traffic control signal is known to operate in accordance with a
predetermined cycle plan, the time period may correspond to a time
slot used in the cycle plan, e.g. over which the cycle times are
constant.
[0123] Of course, it is not necessary that the duration data is in
the form of lists of durations for the different phases in the
given time period, although this is a particularly simple form of
the data that may be easily processed in accordance with the
invention. Any data directly or indirectly indicative of the
durations of the different instances of the red and green phases in
the time period may be used.
[0124] In addition to the duration lists for the red and green
phases, data indicative of a current state of the traffic control
signal is also determined. This data is indicative of the current
phase of the signal, and the time that the signal has had that
phase, i.e. the time elapsed since the transition of the signal to
the current phase. Other forms of timing data may instead to be
used to provide a time reference for the duration data. For
example, a start time for the time period over which duration data
is obtained may be used.
[0125] Once the duration data, i.e. lists of duration times, and
the data relating to the current state of the traffic control
signal, has been determined, the data is used in obtaining data
indicative of a distribution of the times of each phase in at least
a portion of the given time period to which the duration data
relates-step 2 of the FIG. 1 process. It will be appreciated that
the distribution data may be obtained for the same time period to
which the duration data relates, or a portion thereof, to
facilitate processing. The distribution data obtained is preferably
an empirical distribution of the durations for each phase, although
in other arrangements, it is envisaged that instead a model
distribution, e.g. a normal distribution, profile may be fitted to
the data.
[0126] Once the distribution data has been obtained, it is used to
provide data indicative of a probability that the traffic control
signal is green at any future time over a given time period of
interest (step 3 of the FIG. 1 process). It will be appreciated
that this step may not be distinct from the step of obtaining the
distribution data. In addition, an expected waiting time, being the
time that a vehicle arriving at the traffic signal at a particular
time, t, is expected to wait for a green signal may be provided.
The determination of the green signal probability and the expected
waiting time is step 4 of the FIG. 1 process.
[0127] Thus, in preferred embodiments, at least the following data
is obtained as an output of the method based on the duration
lists:
[0128] p(t), the probability for a green signal at time t; and
[0129] W(t), the expected waiting time until green signal at time
t.
[0130] Additionally/alternatively the probability for a red signal
at time t may be obtained as an output. The model for traffic light
phase prediction used in embodiments of the present invention
models the traffic light as a cycle with one green time and one red
time. A yellow phase may be incorporated in either of the red or
green phases depending on the phase sequence of the lights. In the
present examples, the yellow phase is incorporated in the green
phase, which may be suitable for traffic light signals in the
Netherlands and other countries where passage of vehicles is
permitted under the yellow phase. Either or both of the red and
green times may be variable. Where both red and green times are
variable, each may be treated as an independent random
variable.
[0131] The present invention uses knowledge of the distribution of
red and green times to obtain information relating to the
probability that a particular time falls within, for example, a
green phase of a signal.
[0132] In embodiments, the probability that the light is green at
time t, p(t), can be determined by summing the probability of the
signal having a green phase at a time t over all possible cycles.
For instance, in an exemplary embodiment of the present invention,
calculating the probability of a green signal at a time t requires
summing over at least the first, second, third, etc cycles. The
green phase of these cycles can be considered to extend from:
[0133] G.sub.1s to G.sub.1s+G.sub.1
[0134] G.sub.2s to G.sub.2s+G.sub.2
[0135] G.sub.3s to G.sub.3s+G.sub.3, etc.
[0136] with G.sub.2s=G.sub.1s+G.sub.1+R.sub.1, and so on.
[0137] The probability of the light being green at a time t can
then be calculated by considering the probability that a time t
falls within any of these ranges, i.e. by summing over all possible
cycles.
[0138] In embodiments of the present invention, the start of the
ith green time, G.sub.is, the green times, G.sub.i, and the red
times, R.sub.i, may each have an associated probability
distribution function. From these the probability of the green
phase for each cycle extending between any two points in time can
be determined. Using the probability distributions of G.sub.is,
G.sub.i and R.sub.i, the probability of each cycle having a green
phase at a time t can be calculated. The probability of the traffic
control signal having a green phase at a time t can then be
obtained by summing over all possible cycles. These steps may
essentially performed in a single summation.
[0139] As described further below, the probability distributions of
G.sub.i can be obtained from, for example, a list of duration
times. The start of the ith green time may also be obtained from
this data or by using the probability distributions of the green
and red times. In the latter case, the start of the ith green time
may be distributed as:
G is .about. j = 1 i - 1 G j + j = 1 i - 1 + r R j - a = j = 1 i -
1 C j + rR i - a , ##EQU00001##
where G.sub.j and R.sub.j are the stochastic green and red times
and a is the current state of the signal at time t=0.
[0140] The expected waiting time can also be calculated using the
probability data obtained using the duration data. The expected
waiting time can be defined as the expected difference between t
and G.sub.is given that G.sub.is is the start time of the next
green phase.
[0141] In a preferred embodiment, the expected waiting time is
calculated by considering conditional probabilities. For instance,
the expected waiting time at time t may be determined as the sum
of: the expected waiting time given that the signal is red on
arrival multiplied by the probability that the signal is red at
time t; and the expected waiting time given that the signal is
green on arrival multiplied by the probability that the signal is
green at time t.
[0142] The expected waiting time given that the signal is red on
arrival may be calculated using the following equation:
E [ G is - t G is < t + R i and G is > t ] = { E [ R i ] - a
- t if i = 1 , .intg. y = 0 .infin. ( .intg. x = t + a x = t + a +
y xg i - 1 ( x ) dx ) r ( y ) dy .intg. y = 0 .infin. ( .intg. x =
t + a t + a + y g i - 1 ( x ) dx ) r ( y ) dy - a - t if i > 1.
##EQU00002##
[0143] wherein r is the probability function of R.sub.j and [0144]
g.sub.i is the probability density of
[0144] j = 1 i G j + j = 1 i + r R j , i = 1 , 2 , ##EQU00003##
[0145] In the simplest case, it can be assumed that the queue time
is zero. In this case the expected waiting time at a green light is
zero. If the queue time is not zero, then a delay will be
introduced. Methods for incorporating delays to account for queue
times will be discussed later.
[0146] The above equations describe the case where both the red and
green times are variable. If, for example, the red time is static,
then the stochastic red times R.sub.i can be replaced by the
deterministic red time R.
[0147] It may be desirable to increase the calculation speed of
predictions in certain situations, e.g. where the calculations are
to be performed by a navigation device. In such situations various
approximations may be used. For instance, the double integrals
above may be replaced with the following approximation using the
expected value of the red time in the boundaries:
E [ G is - t G is < t + R i and G is > t ] .apprxeq. { E [ R
i ] - a - t if i = 1 , .intg. t + a t + a + E [ R j ] xg i - 1 ( x
) dx .intg. t + a t + a + E [ R j ] g i - 1 ( x ) dx - a - t if i
> 1. ##EQU00004##
[0148] In an exemplary implementation of the present invention, to
increase the calculation speed discretization may be used. In this
case, the above integrals may be replaced with discrete sums. In
one example, the green times, red times and time line in seconds,
and consequently the associated distribution and probability data,
are all discretized.
[0149] A preferred implementation, which is particularly suitable
for use by navigation devices, is as follows.
[0150] Firstly, an exemplary process for calculating the empirical
distributions of the green and red times will be described. The
input to this process may be in the form of a list of red and green
times (taking the yellow phase as part of the green). The start
time of the period of measurement is also known, providing timing
information for the phase data. In other words, in this example, it
is known that the first green time commenced at 7:31:41 in the
morning. The end time of the time period is also known.
[0151] An exemplary list of such times may be as follows. These are
the green times during morning rush hour at a particular traffic
control signal. [0152] 43, 34, 27, 33, 26, 31, 31, 35, 40, 36, 44,
30, 33, 29, 26, [0153] 31, 32, 46, 43, 28, 40, 26, 38, 28, 37, 37,
35, 30, 33, 36. Between these green times, the following red times
have occurred; [0154] 57, 63, 59, 65, 46, 65, 70, 66, 59, 69, 69,
72, 71, 69, 51, [0155] 68, 58, 72, 64, 60, 60, 57, 66, 58, 57, 52,
62, 70, 65, 64. Let x.sub.1, x.sub.2, . . . , x.sub.n be the green
times of the traffic light and y.sub.1, y.sub.2, . . . , y.sub.m
the red times (which are all positive). The main idea of the
implementation of the prediction model is described below. Define
G(t) as the empirical distribution function of the green time and
R(t) of the red time. The empirical distribution functions can be
calculated by:
[0155] G ( t ) = 1 n i = 1 n 1 { x i .ltoreq. t } , t = 0 , 1 ,
##EQU00005## R ( t ) = 1 n i = 1 m 1 { y i .ltoreq. t } , t = 0 , 1
, ##EQU00005.2##
Here 1 is the indicator function. So G(t) is defined as the
fraction of green times that are smaller or equal to t.
[0156] A way in which green times and red times may be used to
obtain distribution data for the durations of the respective phases
is illustrated by reference to FIGS. 2A-C. FIG. 2A illustrates the
measured green times and red times at a particular signal over the
time period 15:00 to 19:00. Data of this type may be used to obtain
the empirical distribution function. FIG. 2B shows histograms
obtained using the green times and red times shown in FIG. 2A for
the specific evening rush hour period of 16:00 to 18:00. FIG. 2C
shows histograms of green times and red times for the same
intersection for the time period 9:30 to 14:30. It can be seen that
outside rush hour the times look normally distributed. During rush
hour at least the red times are not normally distributed.
[0157] Once the distribution data has been obtained, it is used to
provide data indicative of a probability that the traffic control
signal is green at any future time over a given time period of
interest--step 4 of FIG. 1. It will be appreciated that this step
may not be distinct from the step of obtaining the distribution
data.
[0158] In the preferred embodiment, this is done by converting the
empirical distribution for the red and green times to data
indicative of the probability of the traffic signal being green
with respect to time. In order to do this, the green and red time
distributions are converted to probability mass functions. The
probability mass functions can be derived by:
g(t)=G(t)-G(t-1), t=1,2, . . . .
r(t)=R(t)-R(t-1), t=1,2, . . . .
[0159] Using convolution, it is possible to calculate the
probability mass function of the sum of two random variables. The
red and green times can be considered as such variables where the
signal has variable red and green times.
[0160] Thus, the probability mass function of the traffic control
signal cycle (the convolution of the green and red time) can be
calculated as:
{ ( g * r ) ( t ) = i = 0 t g ( i ) r ( t - 1 ) , g min + r min
.ltoreq. t .ltoreq. g max + r max 0 , 0 .ltoreq. t < g min + r
min or t > g max + r max ##EQU00006##
[0161] Next all probability mass functions of necessary cycle
combinations are calculated by using convolutions repeatedly. These
cycle combinations are given by the sums of random variables in the
equations. The required combinations are 1, 2, 3, . . . , max
cycles or 1, 2, 3, . . . , max cycles plus one red time.
[0162] In theory the sums and integrals presented above should be
performed from 1 to infinity. In practice, a good approximation can
be retrieved by summing over only a few terms. In the preferred
implementation, the sums are bounded, otherwise terms will be
divided by numbers which are nearly zero.
[0163] The bounds of the sums can be derived by using the minimum
and maximum green/red times of the given lists, i.e. the green and
red time lists. Let g.sub.min=min (x.sub.i) and r.sub.min=min
(y.sub.i). Define the maximum times as g.sub.max and r.sub.max.
Also use the fact that the green and red times are always
positive.
[0164] By using these convolutions, the probability mass function
for each cycle can be determined. The probability mass functions
are then converted back to distribution functions, using the
following relationship. Let m(t) be a probability mass function,
the corresponding distribution function M(t) can be derived by:
M(t)=.SIGMA..sub.i=1.sup.tm(t), t=1,2, . . . .
[0165] As discussed above, the probability of a green signal at any
moment in the future may be determined by summing over the relevant
distribution functions for each possible cycle. This may be
facilitated, particularly when implemented on a navigation device,
by using the convolutions discussed above.
[0166] Similar convolutions are also used to calculate conditional
expectations, which are used for the expected waiting time until
next green signal as discussed above.
[0167] For the predictions, either discrete or continuous variables
can be used. This discrete case is described by the equations and
may be preferred for implementation on a navigation device. For the
continuous case, the sums become integrals and the probability mass
function becomes a probability density function.
[0168] FIG. 3 illustrates a plot of the probability of the traffic
control signal being green with respect to time over a given future
time period of interest of the type that may be derived in
accordance with the invention. It will be seen that the probability
has repeated minima and maxima, corresponding to the minimum and
maximum probabilities of a green signal in successive cycles of the
signal.
[0169] This data may be used in various manners. One useful
implementation is in obtaining a speed recommendation for a vehicle
approaching the traffic signal--step 5 of the FIG. 1 process.
[0170] In order to derive a speed recommendation, the maxima and
minima of a plot of the type shown in FIG. 3 may be used to
determine time windows reflecting the ranges of time in which a
vehicle arriving at the signal is expected to have the greatest
chance of coinciding with a green signal. The relevant time windows
can be determined by reference to the positions of the maxima, i.e.
being a time window on either side of the central "maximum" time.
Corresponding time windows may be determined around the minima
being indicative of times of arrival when a vehicle has least
chance of coinciding with a green signal (or, conversely, most
chance of coinciding with a red signal). In between these times
there may be time periods when it is less certain as to the phase
of the signal.
[0171] In one exemplary embodiment, these time windows can be
displayed graphically in a manner that helps to visualize the time
periods which are most and least likely to coincide with a green
signal, and if desired, the uncertain time periods therebetween.
One such graphical representation is shown in the lower part of
FIG. 3, and is in the form of a bar 10 with repeating strips of
black, white (with dots) and grey as time increases. For the
avoidance of confusion, this bar may be referred to as the "time"
bar. The black bars correspond to the time windows of greatest
likelihood of a green signal (i.e. those times around the maxima).
The grey bars correspond to time windows with least likelihood of a
green signal (i.e. around the minima). The dotted white bars are
the "uncertain" time windows in between the black and grey
windows.
[0172] The time windows as illustrated by the time bar in FIG. 3
can then be used to derive speed recommendations to enable a
vehicle to arrive at the traffic control signal at a time
corresponding to a green signal. It will be appreciated that such
recommendations may be derived without obtaining the bar 10 shown
in FIG. 3, i.e. direct from the plot of probability against time,
or from other probability data, or even the underlying
duration/distribution data, without calculating a plot as shown in
FIG. 3. It will be appreciated that the time windows indicated by
the time bar of FIG. 3 can readily be converted to speed
recommendation windows for a vehicle approaching the traffic
control signal. These may be displayed to a driver, e.g. using a
PND of the vehicle. The way in which the time windows can be used
to obtain a speed recommendation window will readily be understood.
A suitable conversion may be based upon the time windows
corresponding to the different portions of the bar, and data
indicative of a distance between the vehicle and the traffic
control signal. A time may be determined that the vehicle would
arrive at the signal if it drove with the maximum permissible
speed. This will define the earliest time at which the vehicle
might arrive at the signal. From this time onward, the times
associated with the start and end points of each portion in the bar
shown in FIG. 3 may be derived. It will be appreciated that the
different portions repeat over time, as the traffic control signal
goes through successive cycles. Thus, there may be more than one
range of speed that will correspond to arriving at a "black" time.
A minimum speed of travel may be used to set a maximum time that is
considered, and only speed recommendations above that speed of
travel displayed.
[0173] An exemplary speed recommendation bar 30 is shown in FIG. 4,
together with an indication of how this is derived from the time
colour bar shown in FIG. 3. The vertical bar 20 in FIG. 4 is
indicative of the time bar obtained in FIG. 3. Various lines
emanate from a particular point (the current time) to the left of
the vertical time bar 20 in FIG. 4 and intersect the vertical time
colour bar at different points, i.e. times. Each of these lines is
associated with a different speed of travel, starting with the
maximum permissible speed 35 mph. The trajectory of the line
connects the current time with the time of expected arrival at the
traffic control signal along the time bar, based upon a particular
speed of travel, and taking into account the distance between the
current position of the vehicle and the position of the traffic
control signal.
[0174] Here it can be seen that if the driver travels at 35 mph, he
will arrive at the signal at a time within a black time window.
This is the earliest permissible time of arrival. The speeds of
travel required to arrive at the signal to coincide with the
transition between this black portion of the time bar shown in FIG.
4 and the next portion, a grey one, and then each subsequent
transition between portions is then determined. These speeds are
indicated by the further lines connecting the current time with
each transition in the time bar in FIG. 4, being 24 mph, 14 mph, 11
mph and 8 mph respectively. Speeds below 8 mph are not shown, but
could similarly be derived. In general a minimum speed will usually
be set, and times of arrival associated with speeds below the
minimum speed threshold not considered.
[0175] A speed recommendation bar 30 is then derived based on this
information. As shown in the right hand side of FIG. 4, this
indicates a black portion, being from 24-35 mph. This is a speed
window which should result in arrival within the black time window,
and is therefore a recommended speed window. Below this, between 14
and 24 mph, there is a grey portion in the speed colour bar,
indicating this is not a recommended speed of travel, (which is
expected to result in arrival coinciding with a grey time window,
e.g. a red signal phase). A white (with dots) portion is then
present in the speed bar corresponding to the "uncertain" period,
before another black portion appears. As shown in the screen
display 40 in FIG. 4, a portion of the determined speed
recommendation bar may be displayed to a driver by a PND. The PND
may show only speeds above a certain minimum speed. This may enable
the driver to select a suitable speed to arrive at the traffic
control signal with maximum chance to coincide with a green signal,
by choosing a speed within the black portion of the recommended
speed bar.
[0176] An exemplary method for obtaining the time windows (upon
which the time bar is based) will now be described.
[0177] First the probabilities p(t) are used to obtain black,
dotted white and grey time windows. The probabilities for green
signal are calculated for 1.ltoreq.t.ltoreq.endTime. The endTime
will depend on the minimum speed we want to advise. If the vehicle
arrives at the traffic signal at time t.sub.minspeed with driving
at minimum speed, we can take endTime=t.sub.minspeed. The black
intervals indicate that the probability of green signal is high, so
we try to lead the driver to this region to have maximum
probability to catch green signal. The grey area indicates that the
probability for red signal is high and in the dotted white windows
the predictions are insecure. The middles of the black windows lie
at the local maximums of the probability plot. The middles of the
grey windows lie at the local minimums. The sizes of the black and
grey windows can be changed. This may influence the performance of
the predictions.
[0178] To get the local maximums and minimums, we make use of the
finite difference coefficients. Define the difference as:
diff(t):=p(t+1)-p(t), t=1,2,3, . . . .
We have a local maximum at t.sub.max if:
diff(t.sub.max-1)>0 and diff(t.sub.max+1)<0.
[0179] To be sure that the found local maximum is the local maximum
of the entire green period, we calculate the maximum value of the
neighbourhood. Define locMax(t) as:
loc Max ( t ) : = max i p ( i ) , t - averageGreen * 0.5 .ltoreq. i
.ltoreq. t + averageGreen * 0.5 ##EQU00007##
So we only pick t.sub.max as local maximum if
p(t.sub.max)=locMax(t.sub.max). For an extra check we also want
that p(t.sub.max) is greater than the expected probability:
p(t.sub.max)>averageGreen/(averageGreen+averageRed).
We have a local minimum at t.sub.min if:
diff(t.sub.min-1)<0 and diff(t.sub.min+1)>0.
To be sure that the found local minimum is the local minimum of the
entire red period, we calculate the minimum value of the
neighbourhood. Define the locMin(t.sub.min) as:
loc Min ( t ) : = min i p ( i ) , t - averageRed * 0.5 .ltoreq. i
.ltoreq. t + averageRed * 0.5 ##EQU00008##
So we only pick t.sub.min as local minimum if
p(t.sub.min)=locMin(t.sub.min). For an extra check we also want
that p(t.sub.min) is smaller than the expected probability:
p(t.sub.min)<averageGreen/(averageGreen+averageRed).
If we visualize GI % of the average green time and t.sub.max is a
local maximum. We visualize the values
t.sub.max-averageGreen*(GI/100)/2.ltoreq.t.ltoreq.t.sub.max+averageGreen-
*(GI/100)/2
as black. If we visualize RI % of the average red time and
t.sub.min is a local minimum. We visualize the values
t.sub.min-averageRed*(RI/100)/2.ltoreq.t.ltoreq.t.sub.min+averageRed*(RI-
/100)/2
as grey.
[0180] In case that the traffic light is statically managed, or the
predictions are still constant at the beginning, we visualize black
if:
p(t)>0.99
(if the probability for green light is more than 99% we will
visualize it, this can also be another value). We visualize grey
if:
p(t)<0.01
The times which are not visualized by black or grey, will be
visualized by dotted white. This indicates that the traffic light
in unpredictable at these times.
[0181] When approaching the traffic control signal, speed
recommendations based on the time bar may be displayed in the
vehicle by a PND. By using the distance from the vehicle to the
traffic signal, the time window may be converted to a speed advice
window. First the point in the time window at which the vehicle is
expected to arrive if the vehicle can drive with maximum speed is
determined. This is done using knowledge of the speed limit on the
road towards the intersection where the traffic control signal is
found. Then all times where the section changes in the time bar are
determined. For each of these times a corresponding speed to arrive
exactly at this moment is determined. The determined speeds can be
used to derive a speed recommendation bar, with portions indicating
recommended and not recommended speeds, similar to the time bar.
Note that if the time prediction window is infinite, the speed
recommendation intervals convert to zero. It is clearly not
desirable to recommend excessively low speeds (e.g. 1 mph).
Therefore a minimum speed can be implemented and only speed advices
between the boundaries shall be visualized as shown in FIG. 4.
[0182] As described above, the present invention may also determine
an expected waiting time for the next green signal phase at
different times in a future time period of interest based upon the
phase duration distribution data. This may be determined using the
distribution data, preferably based upon probability data for the
signal having a given phase derived using the distribution
data.
[0183] FIG. 5A-C illustrate plots of expected waiting time against
arrival time at two different traffic control signals which were
obtained using the methods of the present invention. It will be
seen that the waiting time exhibits maxima and minima in a similar
manner to the probability of a green phase for future times. This
is to be expected, given the cyclical operation of the signal. When
arriving at the start of a red phase, the waiting time to a green
signal will be greatest, while at other times a vehicle may arrive
during a green phase, corresponding to no waiting time.
[0184] The determined expected waiting time at a traffic control
signal for future times may be used in several applications.
[0185] In one application, the expected waiting time is used to
provide more accurate estimates as to travel times along a route to
a destination.
[0186] When determining the duration of a route involving one or
more traffic signals, prior art techniques tended to simply add a
delay value for each traffic signal involved to the route duration
obtained by consideration, for example, of average travel speed
data associated with road segments making up the route. This delay
value would be based upon an average delay expected at a traffic
control signal. However, the actual delay at a traffic control
signal will depend upon the phase of the signal encountered by a
vehicle when arriving at a traffic signal, and how long the signal
has had that phase. In accordance with the invention, the
determined expected waiting time may be used to provide a time
delay associated with a particular traffic control signal along a
route which more accurately reflects the actual delay that the
vehicle will experience when arriving at the signal.
[0187] The use of expected waiting time to estimate the expected
time delay associated with traversing intersections along a route
improves the existing route planning and can lead to faster routes
and can also provide the ability to give drivers speed
recommendations to enable them to ride a "green wave" through
multiple sets of signals.
[0188] In one embodiment, data of the type shown in FIGS. 5A-C may
be used to provide a speed recommendation to a vehicle which will
minimise waiting time for a green signal at a traffic control
signal. A speed recommendation may be determined to result in a
vehicle arriving at the traffic signal in one of the periods where
the expected waiting time is at a minimum.
[0189] While the techniques of the present invention have been
described by reference to a single traffic control signal,
corresponding expected waiting time data (and/or, in embodiments,
data indicative of the probability of encountering a green phase)
with respect to time may be obtained for multiple sets of traffic
control signals in a region of interest. For example, p(t) and W(t)
may be obtained for each traffic control signal in a region of
interest, or, alternatively for each traffic control signal along a
pre-calculated route. The latter arrangement may be more efficient
in terms of processing and/or storage requirements.
[0190] Once a route from an origin or current position to a
destination has been generated, the expected waiting time data for
traffic control signals along the route may be used to provide a
more accurate estimate as to the duration of the route.
[0191] The route may be modeled as including various intersections
along its length as shown in FIG. 6. The transit times for portions
of the route between intersections may readily be determined, for
example, by a navigation device as known in the art. This may be
based, for example, upon transit time data associated with road
segments forming the route, or average speed data and length data
associated with the segments, etc. The transit time data may be an
expected transit time based upon historical data, and alternatively
or additionally may take into account "live" traffic conditions,
e.g. actual congestion, etc. The duration of the route may be
obtained by adding the time delay that can be expected to be
encountered at each intersection to the transit times for the
portions of the route between intersections. The time delay at an
intersection is based on the expected waiting time for a vehicle
arriving at the relevant time, i.e. the time that the vehicle is
expected to reach that point along the route.
[0192] Referring to FIG. 6, for example, it takes T.sub.1 seconds
to drive to intersection 1, T.sub.2 seconds to get from
intersection 1 to intersection 2, etc. At each intersection the
vehicle has to wait some time until the next green time. This delay
is estimated by W.sub.i(t). If a route has I intersections with
traffic signals, it can be broken down into I+1 sub routes as
illustrated below:
[0193] Define E[S(i)] as the expected travel time after i sub
routes, just after Ti (for i=1, 2, . . . , I+1). The expected total
travel time from the route (also called the expected sojourn time)
can now be computed with the following recursion:
E[S(1)]=T.sub.1
E[S(2)]=E[S(1)]+E[W.sub.1(E[S(1)])]+T.sub.2
E[S(3)]=E[S(2)]+E[W.sub.2(E[S(2)])]+T.sub.3
. . .
E[S(I+1)]=E[S(I)]+E[W.sub.I(E[S(I)])]+T.sub.I+1
[0194] The expected sojourn time of the route is E[S(I+1)].
[0195] We can allow variation in the travel times between
intersections. If we assume that T.sub.i is uniformly distributed
between a.sub.i and b.sub.i, then we calculate the expected travel
times of the route for T.sub.i=a.sub.i, T.sub.i=(a.sub.i+b.sub.i)/2
and Ti=b.sub.i. If the route has I intersections, the approximation
of the expected sojourn time is the average of all 3.sup.I+1
combinations. If the travel times between intersections are
distributed differently, we can add more weight to the values that
are more likely. It is believed that the distribution of travel
times is dependent upon traffic intensity. The intention of
allowing variation in the travel times between intersections, is to
recognize the situation where it becomes uncertain whether the car
will catch the green signal.
[0196] An example of the determination of a travel time for a route
including intersections having traffic control signals associated
therewith will now be given. FIG. 7 illustrates one exemplary route
in Portland.
[0197] In the example of FIG. 7, we want to calculate the expected
sojourn time along a route including three signalized
intersections. These are the intersections 4115, 4114, and 4113.
The expected waiting times for the traffic control signals at each
of these intersections with respect to arrival time is shown in
FIGS. 5A, 5B and 5C respectively. The route extends from the moment
of arrival at intersection 4115 until leaving intersection 4113.
E[S(4)]-E[S(1)] is calculated, where we define T.sub.4=0 and
T.sub.2=T.sub.3 uniformly between 21 and 25.
[0198] To see how the expected travel times behave, we calculate
the expected sojourn time for each arrival time at the first
intersection. The outcome is illustrated in FIG. 8. In the example
of a route with three traffic signals, we see that the current
states of the traffic signals can make a difference of almost one
minute for the total travel time. At the local minima, the driver
will likely have a green wave.
[0199] By way of explanation, the way in which the sojourn time is
calculated will be explained in greater detail.
[0200] FIGS. 5A, 5B and 5C respectively indicate the expected
waiting time until next green signal at each of the intersections
4115, 4114 and 4113 respectively for t=1, 2, . . . , 200, and are
based on lists of green and red times.
[0201] Let for example the arrival time at intersection 4115 be 50
seconds and T.sub.2=T.sub.3=23. The expected time the vehicle
leaves intersection 4113 will be calculated by:
E[S(1)]=50,
E[S(2)]=50+E[W.sub.1(50)]+23=73,
E[S(3)]=73+E[W.sub.2(73)]+23=109,
E[S(4)]=109+E[W.sub.3(109)]+0=150.
[0202] Thus the expected sojourn time of the route is:
E[S(4)]-E[S(1)]=150-50=100, which is very high because the vehicle
has to wait long at the last intersection.
[0203] For the second example, let the arrival time at intersection
4115 be 25 seconds, T.sub.2=21 and T.sub.3=25. The expected time
the vehicle leaves intersection 4113 will be calculated by:
E[S(1)]=25,
E[S(2)]=25+E[W.sub.1(25)]+21=46,
E[S(3)]=46+E[W.sub.2(46)]+25=71,
E[S(4)]=71+E[W.sub.3(71)]+0=71.
[0204] Hence, the expected sojourn time of the route is:
E[S(4)]-E[S(1)]=71-25=46. In this case, the car does not have to
wait at any intersection. This indicates that the driver will get a
green wave. If a green wave is not possible, we can at least try to
minimize the amount of red signals during a journey. If one of the
expected waiting times is sufficiently small, the traffic signal
phase predictions of the present invention can still guide the
driver through a green wave without giving unnecessarily low
recommended speeds.
[0205] The expected sojourn time for the route shown in FIG. 7 for
all possible arrival times in the range considered is shown in FIG.
9.
[0206] Of course, alternatively or additionally to determining a
travel time for a route, the expected waiting time determined in
accordance with the techniques of the present invention may be
applied in other ways. For example, the expected waiting time may
be used to generate a fastest route from an origin or current
position to a destination, accurately accounting for potential
delay at traffic signals associated with intersections along the
route. A route may be determined that minimises expected waiting
time at signals, whether or not it is a fastest route overall.
Thus, in other embodiments, candidate routes through the road
network may be explored to find a route that minimises expected
waiting time and/or travel time taking into account expected
waiting time.
[0207] The methods of the present invention may be implemented in
various manners: using a server, a navigation device, e.g.
associated with a vehicles, such as a PND, or combinations of both
a server and navigation device.
[0208] With each step, a decision can be taken as to whether to
implement the step at a server or navigation device. The decision
may be based upon the processing and/or storage power available at
a server or navigation device, and how this is to be balanced with
speed of obtaining the result of the processing. In general,
carrying out calculations at a navigation device may be demanding
upon the more limited processing and/or storage capacity of the
device, but will reduce the amount of data that needs to be
transmitted to/from a server, reducing demands upon available
bandwidth.
[0209] Various factors may affect the time and/or processing power
required for carrying out the predictions, e.g. of expected waiting
time or green phase probability. The calculation time of the
prediction increases exponentially if the end time increases. The
calculation time is also larger if the traffic signal is more
dynamically managed or the cycle length is smaller.
[0210] In some embodiments, the server stores data relating to the
durations of the phases of traffic control signals in a geographic
area, e.g. based upon data received from third parties, vehicle
probe data, data received from traffic signals and/or vehicle to
vehicle (V2V) data.
[0211] The relevant duration data may then be sent to a navigation
device when required in relation to a particular traffic signal to
enable the device to calculate the expected waiting time and/or
green signal probability predictions.
[0212] However, other arrangements are possible. The following
table summarises some of the options as to where the specified item
in the left hand column is stored and/or derived. In the table
below "predictions" refers to the prediction of expected waiting
time and/or green and/or red signal probability with respect to
time.
[0213] The items include a database of traffic control signals,
e.g. coordinates of traffic control signals in the region, phase
information for the signals including duration data for different
instances of the phases, i.e. red and green times, and route
calculation.
[0214] The predictions may only be calculated for a next
approaching traffic signal along a route being followed e.g. to
predict the state of an upcoming traffic signal or determine a
speed recommendation to arrive coinciding with a green phase.
However, where a route is being generated, e.g. to minimise waiting
time, or in respect of which a travel time is required taking into
account expected waiting time, it will be necessary to have
knowledge of traffic signals further ahead on the route, and carry
out predictions as to the operation of multiple traffic signals
along the route.
[0215] In the following table some possible options as to where the
various items may be performed are given, with estimated
performance measures.
TABLE-US-00001 TABLE 1 Option 1 Option 2 Option 3 Option 4 Database
traffic Device Server Server Server signals Calculation Device
Server Device Device predictions Calculation Device Server Server
Device route Needed data Traffic signals Fastest route, Fastest
route, traffic Traffic signal data transmission updates traffic
signals signal data e.g. e.g. phase duration predictions phase
duration lists, lists, traffic signal traffic signal updates
updates, Performance Mobile data ++ -- + + CPU device -- ++ + -
Memory device -- ++ + + CPU Server ++ -- - + Memory Server ++ - - -
Score 1 0.5 2.5 1.5
[0216] For example, with option 1, the device, e.g. PND, stores the
necessary phase data for traffic signals, i.e. signals in the
region. This may be in the form of phase duration lists. The device
then calculates the predictions of expected waiting time and/or
probability of green phase with respect to time. The device may
also calculate a route, and be able to obtain a route duration by
determining expected waiting time for traffic signals along the
route, or alternatively may determine a fastest route through the
road network taking into account expected waiting time. In this
arrangements, only updated traffic signal operation data, e.g.
phase data.
[0217] In option 2, the server is arranged to perform all the
operations that the device performed in option 1. The server would
then need to transmit data indicative of a calculated route or
predictions, e.g. expected waiting time, green signal probability,
to a device for use.
[0218] Options 3 and 4 split the various operations between server
and device. Option 3 is seen as most advantageous. In this
arrangement the server stores the traffic signal operation data,
and calculates routes, but predictions are carried out by the
device using traffic signal operation data received from the
server. Similarly the server will transmit route data to the
device.
[0219] By way of example, with an arrangement in accordance with
option 3, if a vehicle is navigating from A to B, a device
associated with the vehicle (e.g. PND) transmits the corresponding
coordinates to the server. The server calculates all possible
routes and transmits the fastest route to the PND.
[0220] The location of B typically will not correspond to an end
destination of the route, but rather may be a suitable location
ahead of the current location, e.g. along an already calculated
part of a route. This avoids the need to obtain and process traffic
signal data for all traffic signals along a long route. If driving
to Rome from Eindhoven for example, there is no need to give
predictions for the traffic signals in Rome at the moment of
leaving Eindhoven. So B can be located on an original route and
will be bounded by distance or estimated travel time (for example:
the upper bound can be 1000 seconds). The device will send the new
locations of A and B repeatedly to check if a faster route is
possible. A may be a current location.
[0221] For the fastest route the server will send the operation
data for the traffic signals along the route, including current
state information and data indicative of the durations of phases,
i.e. phase lists. The server also sends predictions for travel
times between intersections where the signals are located. In
alternative embodiments the estimated travel time between
intersections can also be stored in the device. If the data is more
dynamic, i.e. incorporating "live" traffic information, it may be
more suitably provided by the server. The device can then calculate
the traffic signal phase predictions and/or expected waiting times.
The device may calculate speed recommendations for display to a
driver.
[0222] Every x seconds the device will receive updated traffic
signal information, e.g. current phase, and will then use this
information to update the predictions. As x decreases, the quality
of the prediction improves, but more calculation power and data
transmissions are needed.
[0223] FIG. 10 shows one possible implementation of option 3. On
the left side, an example of a network between A and B is drawn.
The K.sub.ni represents traffic signal i on route n. Define
T.sub.ni as the travel time between intersection i-1 and i for
route n. On the right, we see how option 3 is applied if route n is
chosen as fastest route. Here "probe data" refers to vehicle probe
data, i.e. positional data relating to the movement of devices
associated with vehicles with respect to time, e.g. of mobile
devices, PNDs, etc associated with the vehicles.
[0224] The methods described above may be further refined. In
particular, it is useful to take account of queuing time that may
be experienced by vehicles at traffic control signals. This may
result in additional waiting time being experienced, such that the
driver may in fact miss a green signal even if they arrive at the
traffic signal at a time expected to coincide with a green
signal.
[0225] One way in which compensation may be made for queue time
will now be described.
[0226] With knowledge of the cycle of a traffic signal, i.e. the
green/red signal cycle, it is possible to predict when the queue
length will increase (during red time) and when the queue will
dissolve (during green time). The current arrival rate of vehicles
can be predicted, e.g. using "live" and/or historical sources of
data, e.g. vehicle probe data. For example historical data may be
used based upon historical arrival rates for the relevant timeslot,
and adjusted if this significantly differs from the live situation.
The dissolving rate can also be derived from historical data.
[0227] If the traffic intensity is low enough, the traffic signal
is able to handle all the arriving cars during the green times.
This is illustrated in FIG. 11A, wherein R and G refer to the red
and green phases respectively.
[0228] If the traffic intensity is too high, the traffic signal
becomes saturated and the queue length will increase over time.
This is illustrated in FIG. 11B. This will give rise to significant
predicted delays at the traffic signal, and it may be advised to
avoid the intersection where the traffic signal is located.
[0229] The queue length predictions may be used to derive an
additional time delay attributable to queue time, which can be
incorporated when estimating travel time for a route.
[0230] Any of the methods in accordance with the present invention
may be implemented at least partially using software e.g. computer
programs. The present invention thus also extends to a computer
program comprising computer readable instructions executable to
perform, or to cause a navigation device to perform, a method
according to any of the aspects or embodiments of the invention.
Thus, the invention encompasses a computer program product that,
when executed by one or more processors, cause the one or more
processors to generate suitable images (or other graphical
information) for display on a display screen. The invention
correspondingly extends to a computer software carrier comprising
such software which, when used to operate a system or apparatus
comprising data processing means causes, in conjunction with said
data processing means, said apparatus or system to carry out the
steps of the methods of the present invention. Such a computer
software carrier could be a non-transitory physical storage medium
such as a ROM chip, CD ROM or disk, or could be a signal such as an
electronic signal over wires, an optical signal or a radio signal
such as to a satellite or the like. The present invention provides
a machine readable medium containing instructions which when read
by a machine cause the machine to operate according to the method
of any of the aspects or embodiments of the invention.
[0231] Where not explicitly stated, it will be appreciated that the
invention in any of its aspects may include any or all of the
features described in respect of other aspects or embodiments of
the invention to the extent they are not mutually exclusive. In
particular, while various embodiments of operations have been
described which may be performed in the method and by the
apparatus, it will be appreciated that any one or more or all of
these operations may be performed in the method and by the
apparatus, in any combination, as desired, and as appropriate.
* * * * *